// asmx.c - copyright 1998-2007 Bruce Tomlin #include "asmx.h" #define VERSION_NAME "asmx multi-assembler" //#define ENABLE_REP // uncomment to enable REPEAT pseudo-op (still under development) //#define DOLLAR_SYM // allow symbols to start with '$' (incompatible with $ for hexadecimal constants!) #ifndef VERSION // should be defined on the command line #define VERSION "2.0" #endif #define COPYRIGHT "Copyright 1998-2007 Bruce Tomlin" #define IHEX_SIZE 32 // max number of data bytes per line in hex object file #define MAXSYMLEN 19 // max symbol length (only used in DumpSym()) const int symTabCols = 3; // number of columns for symbol table dump #define MAXMACPARMS 30 // maximum macro parameters #define MAX_INCLUDE 10 // maximum INCLUDE nesting level //#define MAX_BYTSTR 1024 // size of bytStr[] (moved to asmx.h) #define MAX_COND 256 // maximum nesting level of IF blocks #define MAX_MACRO 10 // maximum nesting level of MACRO invocations #if 0 // these should already be defined in sys/types.h (included from stdio.h) typedef unsigned char u_char; typedef unsigned short u_short; typedef unsigned int u_int; typedef unsigned long u_long; #endif #if 0 // moved to asmx.h // a few useful typedefs typedef unsigned char bool; // i guess bool is a c++ thing //enum { FALSE, TRUE }; const bool FALSE = 0; const bool TRUE = 1; typedef char Str255[256]; // generic string type #endif // -------------------------------------------------------------- const char *progname; // pointer to argv[0] struct SymRec { struct SymRec *next; // pointer to next symtab entry u_long value; // symbol value bool defined; // TRUE if defined bool multiDef; // TRUE if multiply defined bool isSet; // TRUE if defined with SET pseudo bool equ; // TRUE if defined with EQU pseudo bool known; // TRUE if value is known char name[1]; // symbol name, storage = 1 + length } *symTab = NULL; // pointer to first entry in symbol table typedef struct SymRec *SymPtr; struct MacroLine { struct MacroLine *next; // pointer to next macro line char text[1]; // macro line, storage = 1 + length }; typedef struct MacroLine *MacroLinePtr; struct MacroParm { struct MacroParm *next; // pointer to next macro parameter name char name[1]; // macro parameter name, storage = 1 + length }; typedef struct MacroParm *MacroParmPtr; struct MacroRec { struct MacroRec *next; // pointer to next macro bool def; // TRUE after macro is defined in pass 2 bool toomany; // TRUE if too many parameters in definition MacroLinePtr text; // macro text MacroParmPtr parms; // macro parms int nparms; // number of macro parameters char name[1]; // macro name, storage = 1 + length } *macroTab = NULL; // pointer to first entry in macro table typedef struct MacroRec *MacroPtr; struct SegRec { struct SegRec *next; // pointer to next segment // bool gen; // FALSE to supress code output (not currently implemented) u_long loc; // locptr for this segment u_long cod; // codptr for this segment char name[1]; // segment name, storage = 1 + length } *segTab = NULL; // pointer to first entry in macro table typedef struct SegRec *SegPtr; #if 0 // moved to asmx.h typedef char OpcdStr[maxOpcdLen+1]; struct OpcdRec { OpcdStr name; // opcode name short typ; // opcode type u_short parm; // opcode parameter }; typedef struct OpcdRec *OpcdPtr; #endif int macroCondLevel; // current IF nesting level inside a macro definition int macUniqueID; // unique ID, incremented per macro invocation int macLevel; // current macro nesting level int macCurrentID[MAX_MACRO]; // current unique ID MacroPtr macPtr[MAX_MACRO]; // current macro in use MacroLinePtr macLine[MAX_MACRO]; // current macro text pointer int numMacParms[MAX_MACRO]; // number of macro parameters Str255 macParmsLine[MAX_MACRO]; // text of current macro parameters char *macParms[MAXMACPARMS * MAX_MACRO]; // pointers to current macro parameters #ifdef ENABLE_REP int macRepeat[MAX_MACRO]; // repeat count for REP pseudo-op #endif struct AsmRec { struct AsmRec *next; // next AsmRec int (*DoCPUOpcode) (int typ, int parm); int (*DoCPULabelOp) (int typ, int parm, char *labl); void (*PassInit) (void); char name[1]; // name of this assembler }; typedef struct AsmRec *AsmPtr; struct CpuRec { struct CpuRec *next; // next CpuRec AsmPtr as; // assembler for CPU type int index; // CPU type index for assembler int endian; // endianness for this CPU int addrWid; // address bus width, ADDR_16 or ADDR_32 int listWid; // listing hex area width, LIST_16 or LIST_24 int wordSize; // addressing word size in bits OpcdPtr opcdTab; // opcdTab[] for this assembler char name[1]; // all-uppercase name of CPU }; typedef struct CpuRec *CpuPtr; // -------------------------------------------------------------- SegPtr curSeg; // current segment SegPtr nullSeg; // default null segment u_long locPtr; // Current program address u_long codPtr; // Current program "real" address int pass; // Current assembler pass bool warnFlag; // TRUE if warning occurred this line bool errFlag; // TRUE if error occurred this line int errCount; // Total number of errors Str255 line; // Current line from input file char *linePtr; // pointer into current line Str255 listLine; // Current listing line bool listLineFF; // TRUE if an FF was in the current listing line bool listFlag; // FALSE to suppress listing source bool listThisLine; // TRUE to force listing this line bool sourceEnd; // TRUE when END pseudo encountered Str255 lastLabl; // last label for '@' temp labels Str255 subrLabl; // current SUBROUTINE label for '.' temp labels bool listMacFlag; // FALSE to suppress showing macro expansions bool macLineFlag; // TRUE if line came from a macro int linenum; // line number in main source file bool expandHexFlag; // TRUE to expand long hex data to multiple listing lines bool symtabFlag; // TRUE to show symbol table in listing bool tempSymFlag; // TRUE to show temp symbols in symbol table listing int condLevel; // current IF nesting level char condState[MAX_COND]; // state of current nesting level enum { condELSE = 1, // ELSE has already been countered at this level condTRUE = 2, // condition is currently true condFAIL = 4 // condition has failed (to handle ELSE after ELSIF) }; int instrLen; // Current instruction length (negative to display as long DB) u_char bytStr[MAX_BYTSTR]; // Current instruction / buffer for long DB statements int hexSpaces; // flags for spaces in hex output for instructions bool showAddr; // TRUE to show LocPtr on listing u_long xferAddr; // Transfer address from END pseudo bool xferFound; // TRUE if xfer addr defined w/ END // Command line parameters Str255 cl_SrcName; // Source file name Str255 cl_ListName; // Listing file name Str255 cl_ObjName; // Object file name bool cl_Err; // TRUE for errors to screen bool cl_Warn; // TRUE for warnings to screen bool cl_List; // TRUE to generate listing file bool cl_Obj; // TRUE to generate object file bool cl_ObjType; // type of object file to generate: enum { OBJ_HEX, OBJ_S9, OBJ_BIN }; // values for cl_Obj u_long cl_Binbase; // base address for OBJ_BIN int cl_S9type; // type of S9 file: 9, 19, 28, or 37 bool cl_Stdout; // TRUE to send object file to stdout bool cl_ListP1; // TRUE to show listing in first assembler pass FILE *source; // source input file FILE *object; // object output file FILE *listing; // listing output file FILE *incbin; // binary include file FILE *(include[MAX_INCLUDE]); // include files Str255 incname[MAX_INCLUDE]; // include file names int incline[MAX_INCLUDE]; // include line number int nInclude; // current include file index bool evalKnown; // TRUE if all operands in Eval were "known" AsmPtr asmTab; // list of all assemblers CpuPtr cpuTab; // list of all CPU types AsmPtr curAsm; // current assembler int curCPU; // current CPU index for current assembler int endian; // CPU endian: UNKNOWN_END, LITTLE_END, BIG_END int addrWid; // CPU address width: ADDR_16, ADDR_32 int listWid; // listing hex area width: LIST_16, LIST_24 int wordSize; // current CPU's addressing size in bits int wordDiv; // scaling factor for current word size int addrMax; // maximum addrWid used OpcdPtr opcdTab; // current CPU's opcode table Str255 defCPU; // default CPU name // -------------------------------------------------------------- enum { // 0x00-0xFF = CPU-specific opcodes // o_Illegal = 0x100, // opcode not found in FindOpcode o_DB = o_Illegal + 1, // DB pseudo-op o_DW, // DW pseudo-op o_DL, // DL pseudo-op o_DWRE, // reverse-endian DW o_DS, // DS pseudo-op o_HEX, // HEX pseudo-op o_FCC, // FCC pseudo-op o_ZSCII, // ZSCII pseudo-op o_ASCIIC, // counted DB pseudo-op o_ASCIIZ, // null-terminated DB pseudo-op o_ALIGN, // ALIGN pseudo-op o_ALIGN_n, // for EVEN pseudo-op o_END, // END pseudo-op o_Include, // INCLUDE pseudo-op o_ENDM, // ENDM pseudo-op #ifdef ENABLE_REP o_REPEND, // REPEND pseudo-op #endif o_MacName, // Macro name o_Processor,// CPU selection pseudo-op // o_LabelOp = 0x1000, // flag to handle opcode in DoLabelOp // 0x1000-0x10FF = CPU-specific label-ops // the following pseudo-ops handle the label field specially // o_EQU = o_LabelOp + 0x100, // EQU and SET pseudo-ops o_ORG = o_EQU + 1, // ORG pseudo-op o_RORG, // RORG pseudo-op o_REND, // REND pseudo-op o_LIST, // LIST pseudo-op o_OPT, // OPT pseudo-op o_ERROR, // ERROR pseudo-op o_MACRO, // MACRO pseudo-op #ifdef ENABLE_REP o_REPEAT, // REPEAT pseudo-op #endif o_Incbin, // INCBIN pseudo-op o_WORDSIZE, // WORDSIZE pseudo-op o_SEG, // SEG pseudo-op o_SUBR, // SUBROUTINE pseudo-op o_IF, // IF pseudo-op o_ELSE, // ELSE pseudo-op o_ELSIF, // ELSIF pseudo-op o_ENDIF // ENDIF pseudo-op }; struct OpcdRec opcdTab2[] = { {"DB", o_DB, 0}, {"FCB", o_DB, 0}, {"BYTE", o_DB, 0}, {"DC.B", o_DB, 0}, {"DFB", o_DB, 0}, {"DEFB", o_DB, 0}, {"DW", o_DW, 0}, {"FDB", o_DW, 0}, {"WORD", o_DW, 0}, {"DC.W", o_DW, 0}, {"DFW", o_DW, 0}, {"DA", o_DW, 0}, {"DEFW", o_DW, 0}, {"DRW", o_DWRE, 0}, {"LONG", o_DL, 0}, {"DL", o_DL, 0}, {"DC.L", o_DL, 0}, {"DS", o_DS, 1}, {"DS.B", o_DS, 1}, {"RMB", o_DS, 1}, {"BLKB", o_DS, 1}, {"DEFS", o_DS, 1}, {"DS.W", o_DS, 2}, {"BLKW", o_DS, 2}, {"DS.L", o_DS, 4}, {"HEX", o_HEX, 0}, {"FCC", o_FCC, 0}, {"ZSCII", o_ZSCII, 0}, {"ASCIIC", o_ASCIIC, 0}, {"ASCIZ", o_ASCIIZ, 0}, {"ASCIIZ", o_ASCIIZ, 0}, {"END", o_END, 0}, {"ENDM", o_ENDM, 0}, {"ALIGN", o_ALIGN, 0}, {"EVEN", o_ALIGN_n, 2}, #ifdef ENABLE_REP {"REPEND", o_REPEND, 0}, #endif {"INCLUDE", o_Include, 0}, {"INCBIN", o_Incbin, 0}, {"PROCESSOR", o_Processor,0}, {"CPU", o_Processor,0}, {"=", o_EQU, 0}, {"EQU", o_EQU, 0}, {":=", o_EQU, 1}, {"SET", o_EQU, 1}, {"DEFL", o_EQU, 1}, {"ORG", o_ORG, 0}, {"AORG", o_ORG, 0}, {"RORG", o_RORG, 0}, {"REND", o_REND, 0}, {"LIST", o_LIST, 0}, {"OPT", o_OPT, 0}, {"ERROR", o_ERROR, 0}, #ifdef ENABLE_REP {"REPEAT", o_REPEAT, 0}, #endif {"MACRO", o_MACRO, 0}, {"SEG", o_SEG, 1}, {"RSEG", o_SEG, 1}, {"SEG.U", o_SEG, 0}, {"SUBR", o_SUBR, 0}, {"SUBROUTINE",o_SUBR, 0}, {"IF", o_IF, 0}, {"ELSE", o_ELSE, 0}, {"ELSIF", o_ELSIF, 0}, {"ENDIF", o_ENDIF, 0}, {"WORDSIZE", o_WORDSIZE, 0}, {"", o_Illegal, 0} }; // -------------------------------------------------------------- #ifdef ENABLE_REP void DoLine(void); // forward declaration #endif // -------------------------------------------------------------- // multi-assembler call vectors int DoCPUOpcode(int typ, int parm) { if (curAsm && curAsm -> DoCPUOpcode) return curAsm -> DoCPUOpcode(typ,parm); else return 0; } int DoCPULabelOp(int typ, int parm, char *labl) { if (curAsm && curAsm -> DoCPULabelOp) return curAsm -> DoCPULabelOp(typ,parm,labl); else return 0; } void PassInit(void) { AsmPtr p; // for each assembler, call PassInit p = asmTab; while(p) { if (p -> PassInit) p -> PassInit(); p = p -> next; } } void *AddAsm(char *name, // assembler name int (*DoCPUOpcode) (int typ, int parm), int (*DoCPULabelOp) (int typ, int parm, char *labl), void (*PassInit) (void) ) { AsmPtr p; p = malloc(sizeof *p + strlen(name)); strcpy(p -> name, name); p -> next = asmTab; p -> DoCPUOpcode = DoCPUOpcode; p -> DoCPULabelOp = DoCPULabelOp; p -> PassInit = PassInit; asmTab = p; return p; } void AddCPU(void *as, // assembler for this CPU char *name, // uppercase name of this CPU int index, // index number for this CPU int endian, // assembler endian int addrWid, // assembler 32-bit int listWid, // listing width int wordSize, // addressing word size in bits struct OpcdRec opcdTab[]) // assembler opcode table { CpuPtr p; p = malloc(sizeof *p + strlen(name)); strcpy(p -> name, name); p -> next = cpuTab; p -> as = (AsmPtr) as; p -> index = index; p -> endian = endian; p -> addrWid = addrWid; p -> listWid = listWid; p -> wordSize = wordSize; p -> opcdTab = opcdTab; cpuTab = p; } CpuPtr FindCPU(char *cpuName) { CpuPtr p; p = cpuTab; while (p) { if (strcmp(cpuName,p->name) == 0) return p; p = p -> next; } return NULL; } void SetWordSize(int wordSize) { wordDiv = (wordSize + 7) / 8; } void CodeFlush(void); // sets up curAsm and curCpu based on cpuName, returns non-zero if success bool SetCPU(char *cpuName) { CpuPtr p; p = FindCPU(cpuName); if (p) { curCPU = p -> index; curAsm = p -> as; endian = p -> endian; addrWid = p -> addrWid; listWid = p -> listWid; wordSize = p -> wordSize; opcdTab = p -> opcdTab; SetWordSize(wordSize); CodeFlush(); // make a visual change in the hex object file return 1; } return 0; } int isalphanum(char c); void Uprcase(char *s); void AsmInit(void) { char *p; #define ASSEMBLER(name) extern void Asm ## name ## Init(void); Asm ## name ## Init(); p = AddAsm("None", NULL, NULL, NULL); AddCPU(p, "NONE", 0, UNKNOWN_END, ADDR_32, LIST_24, 8, NULL); ASSEMBLER(1802); ASSEMBLER(6502); ASSEMBLER(68K); ASSEMBLER(6805); ASSEMBLER(6809); ASSEMBLER(68HC11); ASSEMBLER(68HC16); ASSEMBLER(8048); ASSEMBLER(8051); ASSEMBLER(8085); ASSEMBLER(F8); ASSEMBLER(Jag); ASSEMBLER(Z80); ASSEMBLER(Thumb); ASSEMBLER(ARM); // strcpy(defCPU,"Z80"); // hard-coded default for testing strcpy(line,progname); Uprcase(line); // try to find the CPU name in the executable's name p = line + strlen(line); // start at end of executable name while (p > line && isalphanum(p[-1])) p--; // skip back to last non alpha-numeric character if (!isalphanum(*p)) p++; // advance past last non alpha-numeric character if (p[0] == 'A' && p[1] == 'S' && p[2] == 'M') p = p + 3; // skip leading "ASM" // for each substring, try to find a matching CPU name while (*p) { if (FindCPU(p)) { strcpy(defCPU,p); return; } p++; } } // -------------------------------------------------------------- // error messages /* * Error */ void Error(char *message) { char *name; int line; errFlag = TRUE; errCount++; name = cl_SrcName; line = linenum; if (nInclude >= 0) { name = incname[nInclude]; line = incline[nInclude]; } if (pass == 2) { listThisLine = TRUE; if (cl_List) fprintf(listing, "%s:%d: *** Error: %s ***\n",name,line,message); if (cl_Err) fprintf(stderr, "%s:%d: *** Error: %s ***\n",name,line,message); } } /* * Warning */ void Warning(char *message) { char *name; int line; warnFlag = TRUE; name = cl_SrcName; line = linenum; if (nInclude >= 0) { name = incname[nInclude]; line = incline[nInclude]; } if (pass == 2 && cl_Warn) { listThisLine = TRUE; if (cl_List) fprintf(listing, "%s:%d: *** Warning: %s ***\n",name,line,message); if (cl_Warn) fprintf(stderr, "%s:%d: *** Warning: %s ***\n",name,line,message); } } // -------------------------------------------------------------- // string utilities /* * Debleft deblanks the string s from the left side */ void Debleft(char *s) { char *p = s; while (*p == 9 || *p == ' ') p++; if (p != s) while ((*s++ = *p++)); } /* * Debright deblanks the string s from the right side */ void Debright(char *s) { char *p = s + strlen(s); while (p>s && *--p == ' ') *p = 0; } /* * Deblank removes blanks from both ends of the string s */ void Deblank(char *s) { Debleft(s); Debright(s); } /* * Uprcase converts string s to upper case */ void Uprcase(char *s) { char *p = s; while ((*p = toupper(*p))) p++; } int ishex(char c) { c = toupper(c); return isdigit(c) || ('A' <= c && c <= 'F'); } int isalphaul(char c) { c = toupper(c); return ('A' <= c && c <= 'Z') || c == '_'; } int isalphanum(char c) { c = toupper(c); return isdigit(c) || ('A' <= c && c <= 'Z') || c == '_'; } u_int EvalBin(char *binStr) { u_int binVal; int evalErr; int c; evalErr = FALSE; binVal = 0; while ((c = *binStr++)) { if (c < '0' || c > '1') evalErr = TRUE; else binVal = binVal * 2 + c - '0'; } if (evalErr) { binVal = 0; Error("Invalid binary number"); } return binVal; } u_int EvalOct(char *octStr) { u_int octVal; int evalErr; int c; evalErr = FALSE; octVal = 0; while ((c = *octStr++)) { if (c < '0' || c > '7') evalErr = TRUE; else octVal = octVal * 8 + c - '0'; } if (evalErr) { octVal = 0; Error("Invalid octal number"); } return octVal; } u_int EvalDec(char *decStr) { u_int decVal; int evalErr; int c; evalErr = FALSE; decVal = 0; while ((c = *decStr++)) { if (!isdigit(c)) evalErr = TRUE; else decVal = decVal * 10 + c - '0'; } if (evalErr) { decVal = 0; Error("Invalid decimal number"); } return decVal; } int Hex2Dec(c) { c = toupper(c); if (c > '9') return c - 'A' + 10; return c - '0'; } u_int EvalHex(char *hexStr) { u_int hexVal; int evalErr; int c; evalErr = FALSE; hexVal = 0; while ((c = *hexStr++)) { if (!ishex(c)) evalErr = TRUE; else hexVal = hexVal * 16 + Hex2Dec(c); } if (evalErr) { hexVal = 0; Error("Invalid hexadecimal number"); } return hexVal; } u_int EvalNum(char *word) { int val; // handle C-style 0xnnnn hexadecimal constants if(word[0] == '0') { if (toupper(word[1]) == 'X') { return EvalHex(word+2); } // return EvalOct(word); // 0nnn octal constants are in less demand, though } val = strlen(word) - 1; switch(word[val]) { case 'B': word[val] = 0; val = EvalBin(word); break; case 'O': word[val] = 0; val = EvalOct(word); break; case 'D': word[val] = 0; val = EvalDec(word); break; case 'H': word[val] = 0; val = EvalHex(word); break; default: val = EvalDec(word); break; } return val; } // -------------------------------------------------------------- // listing hex output routines char * ListStr(char *l, char *s) { // copy string to line while (*s) *l++ = *s++; return l; } char * ListByte(char *p, u_char b) { char s[16]; // with extra space for just in case sprintf(s,"%.2X",b); return ListStr(p,s); } char * ListWord(char *p, u_short w) { char s[16]; // with extra space for just in case sprintf(s,"%.4X",w); return ListStr(p,s); } char * ListLong24(char *p, u_long l) { char s[16]; // with extra space for just in case sprintf(s,"%.6lX",l & 0xFFFFFF); return ListStr(p,s); } char * ListLong(char *p, u_long l) { char s[16]; // with extra space for just in case sprintf(s,"%.8lX",l); return ListStr(p,s); } char * ListAddr(char *p,u_long addr) { switch(addrWid) { default: case ADDR_16: p = ListWord(p,addr); break; case ADDR_24: p = ListLong24(p,addr); break; case ADDR_32: // you need listWid == LIST_24 with this too p = ListLong(p,addr); break; }; return p; } char * ListLoc(u_long addr) { char *p; p = ListAddr(listLine,addr); *p++ = ' '; if (listWid == LIST_24 && addrWid == ADDR_16) *p++ = ' '; return p; } // -------------------------------------------------------------- // ZSCII conversion routines u_char zStr[MAX_BYTSTR]; // output data buffer int zLen; // length of output data int zOfs,zPos; // current output offset (in bytes) and bit position int zShift; // current shift lock status (0, 1, 2) char zSpecial[] = "0123456789.,!?_#'\"/\\<-:()"; // special chars table void InitZSCII(void) { zOfs = 0; zPos = 0; zLen = 0; zShift = 0; } void PutZSCII(char nib) { nib = nib & 0x1F; // is it time to start a new word? if (zPos == 3) { // check for overflow if (zOfs >= MAX_BYTSTR) { if (!errFlag) Error("ZSCII string length overflow"); return; } zOfs = zOfs + 2; zPos = 0; } switch(zPos) { case 0: zStr[zOfs] = nib << 2; break; case 1: zStr[zOfs] |= nib >> 3; zStr[zOfs+1] = nib << 5; break; case 2: zStr[zOfs+1] |= nib; break; default: // this shouldn't happen! break; } zPos++; } void PutZSCIIShift(char shift, char nshift) { int lock; lock = 0; // only do a temp shift if next shift is different if (shift == nshift) lock = 2; // generate shift lock if next shift is same switch ((shift - zShift + 3) % 3) { case 0: // no shift break; case 1: // shift forward PutZSCII(0x02 + lock); break; case 2: // shift backwards PutZSCII(0x03 + lock); break; } if (lock) zShift = shift; // update shift lock state } void EndZSCII(void) { // pad final word with shift nibbles while (zPos != 3) PutZSCII(0x05); // set high bit in final word as end-of-text marker zStr[zOfs] |= 0x80; zOfs = zOfs + 2; } int GetZSCIIShift(char ch) { if (ch == ' ') return -2; if (ch == '\n') return -1; if ('a' <= ch && ch <= 'z') return 0; if ('A' <= ch && ch <= 'Z') return 1; return 2; } void ConvertZSCII(void) { // input data is in bytStr[] // input data length is instrLen int i,inpos; // input position char ch; // current and next input byte char *p; // pointer for looking up special chars int shift,nshift; // current and next shift states InitZSCII(); inpos = 0; while (inpos < instrLen) { // get current char and shift ch = bytStr[inpos]; shift = GetZSCIIShift(ch); nshift = shift; // determine next char's shift (skipping blanks and newlines, stopping at end of data) i = inpos + 1; while (i < instrLen && (nshift = GetZSCIIShift(bytStr[i])) < 0) i++; if (i >= instrLen) nshift = zShift; // if end of data, use current shift as "next" shift switch(shift) { case 0: // alpha lower case case 1: // alpha upper case PutZSCIIShift(shift,nshift); PutZSCII(ch - 'A' + 6); break; case 2: // non-alpha if ((p = strchr(zSpecial,ch))) { // numeric and special chars PutZSCIIShift(shift,nshift); PutZSCII(p - zSpecial + 7); } else { // extended char PutZSCIIShift(shift,nshift); PutZSCII(0x06); PutZSCII(ch >> 5); PutZSCII(ch); } break; default: // blank and newline PutZSCII(shift + 2); break; } inpos++; } EndZSCII(); memcpy(bytStr,zStr,zOfs); instrLen = zOfs; } // -------------------------------------------------------------- // token handling // returns 0 for end-of-line, -1 for alpha-numeric, else char value for non-alphanumeric // converts the word to uppercase, too int GetWord(char *word) { u_char c; word[0] = 0; // skip initial whitespace c = *linePtr; while (c == 12 || c == '\t' || c == ' ') c = *++linePtr; // skip comments if (c == ';') while (c) c = *++linePtr; // test for end of line if (c) { // test for alphanumeric token #ifdef DOLLAR_SYM if (isalphanum(c) || (c == '$' && (isalphanum(linePtr[1]) || linePtr[1]=='$'))) #else if (isalphanum(c)) #endif { while (isalphanum(c) || c == '$') { *word++ = toupper(c); c = *++linePtr; } *word = 0; return -1; } else { word[0] = c; word[1] = 0; linePtr++; return c; } } return 0; } // same as GetWord, except it allows '.' chars in alphanumerics and ":=" as a token int GetOpcode(char *word) { u_char c; word[0] = 0; // skip initial whitespace c = *linePtr; while (c == 12 || c == '\t' || c == ' ') c = *++linePtr; // skip comments if (c == ';') while (c) c = *++linePtr; // test for ":=" if (c == ':' && linePtr[1] == '=') { word[0] = ':'; word[1] = '='; word[2] = 0; linePtr = linePtr + 2; return -1; } // test for end of line else if (c) { // test for alphanumeric token if (isalphanum(c) || c=='.') { while (isalphanum(c) || c=='.') { *word++ = toupper(c); c = *++linePtr; } *word = 0; return -1; } else { word[0] = c; word[1] = 0; linePtr++; return c; } } return 0; } void GetFName(char *word) { char *oldLine; int ch; u_char quote; // skip leading whitespace while (*linePtr == ' ' || *linePtr == '\t') linePtr++; oldLine = word; // check for quote at start of file name quote = 0; if (*linePtr == '"' || *linePtr == 0x27) quote = *linePtr++; // continue reading until quote or whitespace or EOL while (*linePtr != 0 && *linePtr != quote && (quote || (*linePtr != ' ' && *linePtr != '\t'))) { ch = *linePtr++; if (ch == '\\' && *linePtr != 0) ch = *linePtr++; *oldLine++ = ch; } *oldLine++ = 0; // if looking for quote, error on end quote if (quote) { if (*linePtr == quote) *linePtr++; else Error("Missing close quote"); } } bool Expect(char *expected) { Str255 s; GetWord(s); if (strcmp(s,expected) != 0) { sprintf(s,"\"%s\" expected",expected); Error(s); return 1; } return 0; } bool Comma() { return Expect(","); } bool RParen() { return Expect(")"); } void EatIt() { Str255 word; while (GetWord(word)); // eat junk at end of line } /* * IllegalOperand */ void IllegalOperand() { Error("Illegal operand"); EatIt(); } /* * MissingOperand */ void MissingOperand() { Error("Missing operand"); EatIt(); } /* * BadMode */ void BadMode() { Error("Illegal addressing mode"); EatIt(); } // find a register name // regList is a space-separated list of register names // FindReg returns: // -2 (aka reg_EOL) if empty string // -1 (aka reg_None) if no register found // 0 if regName is the first register in regList // 1 if regName is the second register in regList // etc. int FindReg(const char *regName, const char *regList) { const char *p; int i; if (!regName[0]) return reg_EOL; i = 0; while (*regList) { p = regName; // compare words while (*p && *p == *regList) { regList++; p++; } // if not match, skip rest of word if (*p || (*regList != 0 && *regList != ' ')) { // skip to next whitespace while (*regList && *regList != ' ') regList++; // skip to next word while (*regList == ' ') regList++; i++; } else return i; } return reg_None; } // get a word and find a register name // regList is a space-separated list of register names // GetReg returns: // -2 (aka reg_EOL) and reports a "missing operand" error if end of line // -1 (aka reg_None) if no register found // 0 if regName is the first register in regList // 1 if regName is the second register in regList // etc. int GetReg(const char *regList) { Str255 word; if (!GetWord(word)) { MissingOperand(); return reg_EOL; } return FindReg(word,regList); } u_int GetBackslashChar(void) { u_char ch; Str255 s; if (*linePtr) { ch = *linePtr++; if (ch == '\\' && *linePtr != 0) // backslash { ch = *linePtr++; switch(ch) { case 'r': ch = '\r'; break; case 'n': ch = '\n'; break; case 't': ch = '\t'; break; case 'x': if (ishex(linePtr[0]) && ishex(linePtr[1])) { s[0] = linePtr[0]; s[1] = linePtr[1]; s[2] = 0; linePtr = linePtr + 2; ch = EvalHex(s); } break; default: break; } } } else ch = -1; return ch; } // -------------------------------------------------------------- // macro handling MacroPtr FindMacro(char *name) { MacroPtr p = macroTab; bool found = FALSE; while (p && !found) { found = (strcmp(p -> name, name) == 0); if (!found) p = p -> next; } return p; } MacroPtr NewMacro(char *name) { MacroPtr p; p = malloc(sizeof *p + strlen(name)); if (p) { strcpy(p -> name, name); p -> def = FALSE; p -> toomany = FALSE; p -> text = NULL; p -> next = macroTab; p -> parms = NULL; p -> nparms = 0; } return p; } MacroPtr AddMacro(char *name) { MacroPtr p; p = NewMacro(name); if (p) macroTab = p; return p; } void AddMacroParm(MacroPtr macro, char *name) { MacroParmPtr parm; MacroParmPtr p; parm = malloc(sizeof *parm + strlen(name)); parm -> next = NULL; strcpy(parm -> name, name); macro -> nparms++; p = macro -> parms; if (p) { while (p -> next) p = p -> next; p -> next = parm; } else macro -> parms = parm; } void AddMacroLine(MacroPtr macro, char *line) { MacroLinePtr m; MacroLinePtr p; m = malloc(sizeof *m + strlen(line)); if (m) { m -> next = NULL; strcpy(m -> text, line); p = macro -> text; if (p) { while (p -> next) p = p -> next; p -> next = m; } else macro -> text = m; } } void GetMacParms(MacroPtr macro) { int i; int n; int quote; char c; char *p; bool done; macCurrentID[macLevel] = macUniqueID++; for (i=0; i=macParms[i + macLevel * MAXMACPARMS] && (*p == ' ' || *p == 9)) *p-- = 0; } if (n > macro -> nparms || n > MAXMACPARMS) Error("Too many macro parameters"); } void DoMacParms() { int i; Str255 word,word2; MacroParmPtr parm; char *p; // pointer to start of word char c; int token; // start at beginning of line linePtr = line; // skip initial whitespace c = *linePtr; while (c == 12 || c == '\t' || c == ' ') c = *++linePtr; // while not end of line p = linePtr; token = GetWord(word); while (token) { // if alphanumeric, search for macro parameter of the same name if (token == -1) { i = 0; parm = macPtr[macLevel] -> parms; while (parm && strcmp(parm -> name, word)) { parm = parm -> next; i++; } // if macro parameter found, replace parameter name with parameter value if (parm) { // copy from linePtr to temp string strcpy(word, linePtr); // copy from corresponding parameter to p strcpy(p, macParms[i + macLevel * MAXMACPARMS]); // point p to end of appended text p = p + strlen(macParms[i + macLevel * MAXMACPARMS]); // copy from temp to p strcpy(p, word); // update linePtr linePtr = p; } } // handle '##' concatenation operator else if (token == '#' && *linePtr == '#') { p = linePtr + 1; // skip second '#' linePtr--; // skip first '#' // skip whitespace to the left while (linePtr > line && linePtr[-1] == ' ') linePtr--; // skip whitespace to the right while (*p == ' ') p++; // copy right side of chopped zone strcpy(word, p); // paste it at new linePtr strcpy(linePtr, word); // and linePtr now even points to where it should } // handle '\0' number of parameters operator else if (token == '\\' && *linePtr == '0') { p = linePtr + 1; // skip '0' linePtr--; // skip '\' // make string of number of parameters sprintf(word2, "%d", numMacParms[macLevel]); // copy right side of chopped zone strcpy(word, p); // paste number strcpy(linePtr, word2); linePtr = linePtr + strlen(word2); // paste right side at new linePtr strcpy(linePtr, word); } // handle '\n' parameter operator else if (token == '\\' && '1' <= *linePtr && *linePtr <= '9') { i = *linePtr - '1'; p = linePtr + 1; // skip 'n' linePtr--; // skip '\' // copy right side of chopped zone strcpy(word, p); // paste parameter strcpy(linePtr, macParms[i + macLevel * MAXMACPARMS]); linePtr = linePtr + strlen(macParms[i + macLevel * MAXMACPARMS]); // paste right side at new linePtr strcpy(linePtr, word); } // handle '\?' unique ID operator else if (token == '\\' && *linePtr == '?') { p = linePtr + 1; // skip '?' linePtr--; // skip '\' // make string of number of parameters sprintf(word2, "%.5d", macCurrentID[macLevel]); // copy right side of chopped zone strcpy(word, p); // paste number strcpy(linePtr, word2); linePtr = linePtr + strlen(word2); // paste right side at new linePtr strcpy(linePtr, word); } /* just use "\##" instead to avoid any confusion with \\ inside of DB pseudo-op // handle '\\' escape else if (token == '\\' && *linePtr == '\\') { p = linePtr + 1; // skip second '\' // copy right side of chopped zone strcpy(word, p); // paste right side at new linePtr strcpy(linePtr, word); } */ // skip initial whitespace c = *linePtr; while (c == 12 || c == '\t' || c == ' ') c = *++linePtr; p = linePtr; token = GetWord(word); } } void DumpMacro(MacroPtr p) { MacroLinePtr line; MacroParmPtr parm; if (cl_List) { fprintf(listing,"--- Macro '%s' ---", p -> name); fprintf(listing," def = %d, nparms = %d\n", p -> def, p -> nparms); // dump parms here fprintf(listing,"Parms:"); for (parm = p->parms; parm; parm = parm->next) { fprintf(listing," '%s'",parm->name); } fprintf(listing,"\n"); // dump text here for (line = p->text; line; line = line->next) fprintf(listing," '%s'\n",line->text); } } void DumpMacroTab(void) { struct MacroRec *p; p = macroTab; while (p) { DumpMacro(p); p = p -> next; } } // -------------------------------------------------------------- // opcodes and symbol table /* * FindOpcodeTab - finds an entry in an opcode table */ // special compare for opcodes to allow "*" wildcard int opcode_strcmp(const char *s1, const char *s2) { while (*s1 == *s2++) if (*s1++ == 0) return 0; if (*s1 == '*') return 0; // this is the magic return (*s1 - *(s2 - 1)); } OpcdPtr FindOpcodeTab(OpcdPtr p, char *name, int *typ, int *parm) { bool found = FALSE; // while (p -> typ != o_Illegal && !found) while (*(p -> name) && !found) { found = (opcode_strcmp(p -> name, name) == 0); if (!found) p++; else { *typ = p -> typ; *parm = p -> parm; } } if (!found) p = NULL; // because this is an array, not a linked list return p; } /* * FindOpcode - finds an opcode in either the generic or CPU-specific * opcode tables, or as a macro name */ OpcdPtr GetFindOpcode(char *opcode, int *typ, int *parm, MacroPtr *macro) { *typ = o_Illegal; *parm = 0; *macro = NULL; OpcdPtr p; int len; p = NULL; if (GetOpcode(opcode)) { if (opcdTab) p = FindOpcodeTab(opcdTab, opcode, typ, parm); if (!p) { if (opcode[0] == '.') opcode++; // allow pseudo-ops to be invoked as ".OP" p = FindOpcodeTab((OpcdPtr) &opcdTab2, opcode, typ, parm); } if (p) { // if wildcard was matched, back up linePtr // NOTE: if wildcard matched an empty string, linePtr will be // unchanged and point to whitespace len = strlen(p->name); if (len && (p->name[len-1] == '*')) { linePtr = linePtr - (strlen(opcode) - len + 1); } } else { if ((*macro = FindMacro(opcode))) { *typ = o_MacName; p = opcdTab2; // return dummy non-null valid opcode pointer } } } if (pass == 2 && !strcmp(opcode,"FROB")) { printf("*** FROB typ=%d, parm=%d, macro=%.8X, p=%.8X\n",*typ,*parm,(int) macro,(int) p); } return p; } /* * FindSym */ SymPtr FindSym(char *symName) { SymPtr p = symTab; bool found = FALSE; while (p && !found) { found = (strcmp(p -> name, symName) == 0); if (!found) p = p -> next; } return p; } /* * AddSym */ SymPtr AddSym(char *symName) { SymPtr p; p = malloc(sizeof *p + strlen(symName)); strcpy(p -> name, symName); p -> value = 0; p -> next = symTab; p -> defined = FALSE; p -> multiDef = FALSE; p -> isSet = FALSE; p -> equ = FALSE; p -> known = FALSE; symTab = p; return p; } /* * RefSym */ int RefSym(char *symName, bool *known) { SymPtr p; int i; Str255 s; if ((p = FindSym(symName))) { if (!p -> defined) { sprintf(s, "Symbol '%s' undefined", symName); Error(s); } switch(pass) { case 1: if (!p -> defined) *known = FALSE; break; case 2: if (!p -> known) *known = FALSE; break; } #if 0 // FIXME: possible fix that may be needed for 16-bit address if (addrWid == ADDR_16) return (short) p -> value; // sign-extend from 16 bits #endif return p -> value; } { // check for 'FFH' style constants here i = strlen(symName)-1; if (toupper(symName[i]) != 'H') i = -1; else while (i>0 && ishex(symName[i-1])) i--; if (i == 0) { strncpy(s, symName, 255); s[strlen(s)-1] = 0; return EvalHex(s); } else { p = AddSym(symName); *known = FALSE; // sprintf(s, "Symbol '%s' undefined", symName); // Error(s); } } return 0; } /* * DefSym */ void DefSym(char *symName, u_long val, bool setSym, bool equSym) { SymPtr p; Str255 s; if (symName[0]) // ignore null string symName { p = FindSym(symName); if (p == NULL) p = AddSym(symName); if (!p -> defined || (p -> isSet && setSym)) { p -> value = val; p -> defined = TRUE; p -> isSet = setSym; p -> equ = equSym; } else if (p -> value != val) { p -> multiDef = TRUE; if (pass == 2 && !p -> known) sprintf(s, "Phase error"); else sprintf(s, "Symbol '%s' multiply defined",symName); Error(s); } if (pass == 0 || pass == 2) p -> known = TRUE; } } void DumpSym(SymPtr p, char *s, int *w) { // // ####### // char *s2; int n,len,max; n = 0; *w = 1; max = MAXSYMLEN; s2 = p->name; len = strlen(s2); while (max-1 < len) { *w = *w + 1; max = max + MAXSYMLEN + 8; // 8 = number of extra chars between symbol names } while(*s2 && n < max) { *s++ = *s2++; n++; } while (n < max) { *s++ = ' '; n++; } switch(addrMax) { default: case ADDR_16: sprintf(s, "%.4lX ", p->value & 0xFFFF); break; case ADDR_24: #if 0 sprintf(s, "%.6lX ", p->value & 0xFFFFFF); break; #endif case ADDR_32: sprintf(s, "%.8lX ", p->value); break; } s = s + strlen(s); n = 0; if (!p -> defined) {*s++ = 'U'; n++;} // Undefined if ( p -> multiDef) {*s++ = 'M'; n++;} // Multiply defined if ( p -> isSet) {*s++ = 'S'; n++;} // Set if ( p -> equ) {*s++ = 'E'; n++;} // Equ while (n < 3) { *s++ = ' '; n++; } *s = 0; } void DumpSymTab(void) { struct SymRec *p; int i,w; Str255 s; i = 0; p = symTab; while (p) { if (tempSymFlag || !(strchr(p->name,'.') || strchr(p->name,'@'))) { DumpSym(p,s,&w); p = p -> next; // force a newline if new symbol won't fit on current line if (i+w > symTabCols) { if (cl_List) fprintf(listing, "\n"); i = 0; } // if last symbol or if symbol fills line, deblank and print it if (p == NULL || i+w >= symTabCols) { Debright(s); if (cl_List) fprintf(listing, "%s\n", s); i = 0; } // otherwise just print it and count its width else { if (cl_List) fprintf(listing, "%s", s); i = i + w; } } else p = p -> next; } } void SortSymTab() { SymPtr i,j; // pointers to current elements SymPtr ip,jp; // pointers to previous elements SymPtr t; // temp for swapping // yes, it's a linked-list bubble sort if (symTab != NULL) { ip = NULL; i = symTab; jp = i; j = i -> next; while (j != NULL) { while (j != NULL) { if (strcmp(i->name,j->name) > 0) // (i->name > j->name) { if (ip != NULL) ip -> next = j; else symTab = j; if (i == jp) { i -> next = j -> next; j -> next = i; } else { jp -> next = i; t = i -> next; i -> next = j -> next; j -> next = t; } t = i; i = j; j = t; } jp = j; j = j -> next; } ip = i; i = i -> next; jp = i; j = i -> next; } } } // -------------------------------------------------------------- // expression evaluation int Eval0(void); // forward declaration int Factor(void) { Str255 word,s; int token; int val; char *oldLine; SymPtr p; token = GetWord(word); val = 0; switch(token) { case 0: MissingOperand(); break; case '%': GetWord(word); val = EvalBin(word); break; case '$': if (ishex(*linePtr)) { GetWord(word); val = EvalHex(word); break; } // fall-through... case '*': val = locPtr; #if 0 // FIXME: possible fix that may be needed for 16-bit address if (addrWid == ADDR_16) val = (short) val; // sign-extend from 16 bits #endif val = val / wordDiv; break; case '+': val = Factor(); break; case '-': val = -Factor(); break; case '~': val = ~Factor(); break; case '!': val = !Factor(); break; case '<': val = Factor() & 0xFF; break; case '>': val = (Factor() >> 8) & 0xFF; break; case '(': val = Eval0(); RParen(); break; case '[': val = Eval0(); Expect("]"); break; case 0x27: // single quote #if 1 // enable multi-char single-quote constants val = 0; while (*linePtr != 0x27 && *linePtr != 0) { val = val * 256 + GetBackslashChar(); } if (*linePtr == 0x27) linePtr++; else Error("Missing close quote"); #else if ((val = GetBackslashChar()) >= 0) { if (*linePtr && *linePtr != 0x27) val = val * 256 + GetBackslashChar(); if (*linePtr == 0x27) linePtr++; else Error("Missing close quote"); } else MissingOperand(); #endif break; case '.': // check for ".." oldLine = linePtr; val = GetWord(word); if (val == '.') { GetWord(word); // check for "..DEF" operator if (strcmp(word,"DEF") == 0) { val = 0; if (GetWord(word) == -1) { p = FindSym(word); val = (p && (p -> known || pass == 1)); } else IllegalOperand(); break; } // check for "..UNDEF" operator if (strcmp(word,"UNDEF") == 0) { val = 0; if (GetWord(word) == -1) { p = FindSym(word); val = !(p && (p -> known || pass == 1)); } else IllegalOperand(); break; } // invalid ".." operator // rewind and return "current location" linePtr = oldLine; break; } // check for '.' as "current location" else if (val != -1) { linePtr = oldLine; val = locPtr; #if 0 // FIXME: possible fix that may be needed for 16-bit address if (addrWid == ADDR_16) val = (short) val; // sign-extend from 16 bits #endif val = val / wordDiv; break; } // now try it as a local ".symbol" linePtr = oldLine; // fall-through... case '@': GetWord(word); if (token == '.' && subrLabl[0]) strcpy(s,subrLabl); else strcpy(s,lastLabl); s[strlen(s)+1] = 0; s[strlen(s)] = token; strcat(s,word); val = RefSym(s, &evalKnown); break; case -1: if ((word[0] == 'H' || word[0] == 'L') && word[1] == 0 && *linePtr == '(') { // handle H() and L() from vintage Atari 7800 source code // note: no whitespace allowed before the open paren! token = word[0]; // save 'H' or 'L' GetWord(word); // eat left paren val = Eval0(); // evaluate sub-expression RParen(); // check for right paren if (token == 'H') val = (val >> 8) & 0xFF; if (token == 'L') val = val & 0xFF; break; } if (isdigit(word[0])) val = EvalNum(word); else val = RefSym(word,&evalKnown); break; default: MissingOperand(); break; } return val; } int Term(void) { Str255 word; int token; int val,val2; char *oldLine; val = Factor(); oldLine = linePtr; token = GetWord(word); while (token == '*' || token == '/' || token == '%') { switch(token) { case '*': val = val * Factor(); break; case '/': val2 = Factor(); if (val2) val = val / val2; else { Warning("Division by zero"); val = 0; } break; case '%': val2 = Factor(); if (val2) val = val % val2; else { Warning("Division by zero"); val = 0; } break; } oldLine = linePtr; token = GetWord(word); } linePtr = oldLine; return val; } int Eval2(void) { Str255 word; int token; int val; char *oldLine; val = Term(); oldLine = linePtr; token = GetWord(word); while (token == '+' || token == '-') { switch(token) { case '+': val = val + Term(); break; case '-': val = val - Term(); break; } oldLine = linePtr; token = GetWord(word); } linePtr = oldLine; return val; } int Eval1(void) { Str255 word; int token; int val; char *oldLine; val = Eval2(); oldLine = linePtr; token = GetWord(word); while ((token == '<' && *linePtr != token) || (token == '>' && *linePtr != token) || token == '=' || (token == '!' && *linePtr == '=')) { switch(token) { case '<': if (*linePtr == '=') {linePtr++; val = (val <= Eval2());} else val = (val < Eval2()); break; case '>': if (*linePtr == '=') {linePtr++; val = (val >= Eval2());} else val = (val > Eval2()); break; case '=': if (*linePtr == '=') linePtr++; // allow either one or two '=' signs val = (val == Eval2()); break; case '!': linePtr++; val = (val != Eval2()); break; } oldLine = linePtr; token = GetWord(word); } linePtr = oldLine; return val; } int Eval0(void) { Str255 word; int token; int val; char *oldLine; val = Eval1(); oldLine = linePtr; token = GetWord(word); while (token == '&' || token == '|' || token == '^' || (token == '<' && *linePtr == '<') || (token == '>' && *linePtr == '>')) { switch(token) { case '&': if (*linePtr == '&') {linePtr++; val = ((val & Eval1()) != 0);} else val = val & Eval1(); break; case '|': if (*linePtr == '|') {linePtr++; val = ((val | Eval1()) != 0);} else val = val | Eval1(); break; case '^': val = val ^ Eval1(); break; case '<': linePtr++; val = val << Eval1(); break; case '>': linePtr++; val = val >> Eval1(); break; } oldLine = linePtr; token = GetWord(word); } linePtr = oldLine; return val; } int Eval(void) { evalKnown = TRUE; return Eval0(); } void CheckByte(int val) { if (!errFlag && (val < -128 || val > 255)) Warning("Byte out of range"); } void CheckStrictByte(int val) { if (!errFlag && (val < -128 || val > 127)) Warning("Byte out of range"); } void CheckWord(int val) { if (!errFlag && (val < -32768 || val > 65535)) Warning("Word out of range"); } void CheckStrictWord(int val) { if (!errFlag && (val < -32768 || val > 32767)) Warning("Word out of range"); } int EvalByte(void) { long val; val = Eval(); CheckByte(val); return val & 0xFF; } int EvalBranch(int instrLen) { long val; val = Eval(); val = val - locPtr - instrLen; if (!errFlag && (val < -128 || val > 127)) Error("Short branch out of range"); return val & 0xFF; } int EvalWBranch(int instrLen) { long val; val = Eval(); val = val - locPtr - instrLen; if (!errFlag && (val < -32768 || val > 32767)) Error("Word branch out of range"); return val; } int EvalLBranch(int instrLen) { long val; val = Eval(); val = val - locPtr - instrLen; return val; } // -------------------------------------------------------------- // object file generation u_char hex_buf[IHEX_SIZE]; // buffer for current line of object data u_long hex_len; // current size of object data buffer u_long hex_base; // address of start of object data buffer u_long hex_addr; // address of next byte in object data buffer u_short hex_page; // high word of address for intel hex file u_long bin_eof; // current end of file when writing binary file // Intel hex format: // // :aabbbbccdddd...ddee // // aa = record data length (the number of dd bytes) // bbbb = address for this record // cc = record type // 00 = data (data is in the dd bytes) // 01 = end of file (bbbb is transfer address) // 02 = extended segment address record // dddd (two bytes) represents the segment address // 03 = Start segment address record // dddd (two bytes) represents the segment of the transfer address // 04 = extended linear address record // dddd (two bytes) represents the high address word // 05 = Start linear address record // dddd (two bytes) represents the high word of the transfer address // dd... = data bytes if record type needs it // ee = checksum byte: add all bytes aa through dd // and subtract from 256 (2's complement negate) void write_ihex(u_long addr, u_char *buf, int len, int rectype) { int i,chksum; // if transfer record with long address, write extended address record if (rectype == 1 && (addr & 0xFFFF0000)) write_ihex(addr >> 16, buf, 0, 5); // if data record with long address, write extended address record if (rectype == 0 && (addr >> 16) != hex_page) { write_ihex(addr >> 16, buf, 0, 4); hex_page = addr >> 16; } // compute initial checksum from length, address, and record type chksum = len + (addr >> 8) + addr + rectype; // print length, address, and record type fprintf(object,":%.2X%.4lX%.2X",len,addr & 0xFFFF,rectype); // print data while updating checksum for (i=0; i
, // 5 number of data records in preceeding block // 6 unused // 7 ending record for S3 records // 8 ending record for S2 records // 9 ending record for S1 records // bb = record data length (from bb through ee) // cc... = address for this record, 2 bytes for S1/S9, 3 bytes for S2/S8, 4 bytes for S3/S7 // dd... = data bytes if record type needs it // ee = checksum byte: add all bytes bb through dd // and subtract from 255 (1's complement) void write_srec(u_long addr, u_char *buf, int len, int rectype) { int i,chksum; // start checksum with length and 16-bit address chksum = len+3 + ((addr >> 8) & 0xFF) + (addr & 0xFF); // determine S9 record type if (rectype) i = cl_S9type % 10; // xfer record = S9/S8/S7 else i = cl_S9type / 10; // code record = S1/S2/S3 // print length and address, and update checksum for long address switch(cl_S9type) { case 37: fprintf(object,"S%d%.2X%.8lX", i, len+5, addr); chksum = chksum + ((addr >> 24) & 0xFF) + ((addr >> 16) & 0xFF) + 2; break; case 28: fprintf(object,"S%d%.2X%.6lX", i, len+4, addr & 0xFFFFFF) + 1; chksum = chksum + ((addr >> 16) & 0xFF); break; default: if (i == 0) i = 1; // handle "-s9" option fprintf(object,"S%d%.2X%.4lX", i, len+3, addr & 0xFFFF); } // print data while updating checksum for (i=0; i cl_Binbase) { buf = buf + cl_Binbase - addr; addr = cl_Binbase; } // if addr is beyond current EOF, write (addr-bin_eof) bytes of 0xFF padding if (addr - cl_Binbase > bin_eof) { fseek(object, bin_eof, SEEK_SET); for (i=0; i < addr - cl_Binbase - bin_eof; i++) fputc(0xFF, object); } // seek to addr and write buf fseek(object, addr - cl_Binbase, SEEK_SET); fwrite(buf, 1, len, object); // update EOF of object file i = ftell(object); if (i > bin_eof) bin_eof = i; //fflush(object); } } // rectype 0 = code, rectype 1 = xfer void write_hex(u_long addr, u_char *buf, int len, int rectype) { if (cl_Obj || cl_Stdout) { switch(cl_ObjType) { default: case OBJ_HEX: write_ihex(addr, buf, len, rectype); break; case OBJ_S9: write_srec(addr, buf, len, rectype); break; case OBJ_BIN: write_bin (addr, buf, len, rectype); break; } } } void CodeInit(void) { hex_len = 0; hex_base = 0; hex_addr = 0; hex_page = 0; bin_eof = 0; } void CodeFlush(void) { if (hex_len) { write_hex(hex_base, hex_buf, hex_len, 0); hex_len = 0; hex_base = hex_base + hex_len; hex_addr = hex_base; } } void CodeOut(int byte) { if (pass == 2) { if (codPtr != hex_addr) { CodeFlush(); hex_base = codPtr; hex_addr = codPtr; } hex_buf[hex_len++] = byte; hex_addr++; if (hex_len == IHEX_SIZE) CodeFlush(); } locPtr++; codPtr++; } void CodeEnd(void) { CodeFlush(); if (pass == 2) { if (xferFound) write_hex(xferAddr, hex_buf, 0, 1); } } void CodeAbsOrg(int addr) { codPtr = addr; locPtr = addr; } void CodeRelOrg(int addr) { locPtr = addr; } void CodeXfer(int addr) { xferAddr = addr; xferFound = TRUE; } void AddLocPtr(int ofs) { codPtr = codPtr + ofs; locPtr = locPtr + ofs; } // -------------------------------------------------------------- // instruction format calls // clear the length of the instruction void InstrClear(void) { instrLen = 0; hexSpaces = 0; } // add a byte to the instruction void InstrAddB(u_char b) { bytStr[instrLen++] = b; hexSpaces |= 1< 255 // this is for opcodes with pre-bytes void InstrAddX(u_long op) { // if ((op & 0xFFFFFF00) == 0) hexSpaces |= 1; // to indent single-byte instructions // if (op & 0xFF000000) bytStr[instrLen++] = op >> 24; // if (op & 0xFFFF0000) bytStr[instrLen++] = op >> 16; if (op & 0xFFFFFF00) bytStr[instrLen++] = op >> 8; bytStr[instrLen++] = op & 255; hexSpaces |= 1<> 8; } else if (endian == BIG_END) { bytStr[instrLen++] = w >> 8; bytStr[instrLen++] = w & 255; } else Error("CPU endian not defined"); hexSpaces |= 1<> 8) & 255; bytStr[instrLen++] = (l >> 16) & 255; } else if (endian == BIG_END) { bytStr[instrLen++] = (l >> 16) & 255; bytStr[instrLen++] = (l >> 8) & 255; bytStr[instrLen++] = l & 255; } else Error("CPU endian not defined"); hexSpaces |= 1<> 8) & 255; bytStr[instrLen++] = (l >> 16) & 255; bytStr[instrLen++] = (l >> 24) & 255; } else if (endian == BIG_END) { bytStr[instrLen++] = (l >> 24) & 255; bytStr[instrLen++] = (l >> 16) & 255; bytStr[instrLen++] = (l >> 8) & 255; bytStr[instrLen++] = l & 255; } else Error("CPU endian not defined"); hexSpaces |= 1< name, name) == 0); if (!found) p = p -> next; } return p; } SegPtr AddSeg(char *name) { SegPtr p; p = malloc(sizeof *p + strlen(name)); p -> next = segTab; // p -> gen = TRUE; p -> loc = 0; p -> cod = 0; strcpy(p -> name, name); segTab = p; return p; } void SwitchSeg(SegPtr seg) { CodeFlush(); curSeg -> cod = codPtr; curSeg -> loc = locPtr; curSeg = seg; codPtr = curSeg -> cod; locPtr = curSeg -> loc; } // -------------------------------------------------------------- // text I/O int OpenInclude(char *fname) { if (nInclude == MAX_INCLUDE - 1) return -1; nInclude++; include[nInclude] = NULL; incline[nInclude] = 0; strcpy(incname[nInclude],fname); include[nInclude] = fopen(fname, "r"); if (include[nInclude]) return 1; nInclude--; return 0; } void CloseInclude(void) { if (nInclude < 0) return; fclose(include[nInclude]); include[nInclude] = NULL; nInclude--; } int ReadLine(FILE *file, char *line, int max) { int c = 0; int len = 0; macLineFlag = TRUE; // if at end of macro and inside a nested macro, pop the stack while (macLevel > 0 && macLine[macLevel] == NULL) { #ifdef ENABLE_REP if (macRepeat[macLevel] > 0) { if (macRepeat[macLevel]--) macLine = macPtr[macLevel] -> text; else { free(macPtr[macLevel]); macLevel--; } } else #endif macLevel--; } // if there is still another macro line to process, get it if (macLine[macLevel] != NULL) { strcpy(line, macLine[macLevel] -> text); macLine[macLevel] = macLine[macLevel] -> next; DoMacParms(); } else { // else we weren't in a macro or we just ran out of macro macLineFlag = FALSE; if (nInclude >= 0) incline[nInclude]++; else linenum++; macPtr[macLevel] = NULL; while (max > 1) { c = fgetc(file); *line = 0; switch(c) { case EOF: if (len == 0) return 0; case '\n': return 1; case '\r': c = fgetc(file); if (c != '\n') { ungetc(c,file); c = '\r'; } return 1; default: *line++ = c; max--; len++; break; } } while (c != EOF && c != '\n') c = fgetc(file); } return 1; } int ReadSourceLine(char *line, int max) { int i; while (nInclude >= 0) { i = ReadLine(include[nInclude], line, max); if (i) return i; CloseInclude(); } return ReadLine(source, line, max); } void ListOut(bool showStdErr) { /* uncomment this block if you want form feeds to be sent to the listing file if (listLineFF && cl_List) fputc(12,listing); */ Debright(listLine); if (cl_List) fprintf(listing,"%s\n",listLine); if (pass == 2 && showStdErr && ((errFlag && cl_Err) || (warnFlag && cl_Warn))) fprintf(stderr,"%s\n",listLine); } void CopyListLine(void) { int n; char c,*p,*q; p = listLine; q = line; listLineFF = FALSE; // the old version // strcpy(listLine, " "); // 16 blanks // strncat(listLine, line, 255-16); if (listWid == LIST_24) for (n=24; n; n--) *p++ = ' '; // 24 blanks at start of line else for (n=16; n; n--) *p++ = ' '; // 16 blanks at start of line while (n < 255-16 && (c = *q++)) // copy rest of line, stripping out form feeds { if (c == 12) listLineFF = TRUE; // if a form feed was found, remember it for later else { *p++ = c; n++; } } *p = 0; // string terminator } // -------------------------------------------------------------- // main assembler loops void DoOpcode(int typ, int parm) { int val; int i,n; Str255 word,s; char *oldLine; int token; int ch; u_char quote; char *p; if (DoCPUOpcode(typ, parm)) return; switch(typ) { case o_ZSCII: case o_ASCIIC: case o_ASCIIZ: case o_DB: instrLen = 0; oldLine = linePtr; token = GetWord(word); if (token == 0) MissingOperand(); if (typ == o_ASCIIC) bytStr[instrLen++] = 0; while (token) { if ((token == '\'' && *linePtr && linePtr[1] != '\'') || token == '"') { quote = token; while (token == quote) { while (*linePtr != 0 && *linePtr != token) { ch = GetBackslashChar(); if ((ch >= 0) && (instrLen < MAX_BYTSTR)) bytStr[instrLen++] = ch; } token = *linePtr; if (token) linePtr++; else Error("Missing close quote"); if (token == quote && *linePtr == quote) // two quotes together { if (instrLen < MAX_BYTSTR) bytStr[instrLen++] = *linePtr++; } else token = *linePtr; } } else { linePtr = oldLine; val = EvalByte(); if (instrLen < MAX_BYTSTR) bytStr[instrLen++] = val; } token = GetWord(word); oldLine = linePtr; if (token == ',') { token = GetWord(word); if (token == 0) MissingOperand(); } else if (token) { linePtr = oldLine; Comma(); token = 0; } else if (errFlag) // this is necessary to keep EvalByte() errors token = 0; // from causing phase errors } if (instrLen >= MAX_BYTSTR || (typ == o_ASCIIC && instrLen > 256)) { Error("String too long"); instrLen = MAX_BYTSTR; } switch(typ) { case o_ASCIIC: if (instrLen > 255) bytStr[0] = 255; else bytStr[0] = instrLen-1; break; case o_ZSCII: ConvertZSCII(); break; case o_ASCIIZ: if (instrLen < MAX_BYTSTR) bytStr[instrLen++] = 0; break; default: break; } instrLen = -instrLen; break; case o_DW: case o_DWRE: // reverse-endian DW if (endian != LITTLE_END && endian != BIG_END) { Error("CPU endian not defined"); break; } instrLen = 0; oldLine = linePtr; token = GetWord(word); if (token == 0) MissingOperand(); while (token) { #if 1 // enable padded string literals if ((token == '\'' && *linePtr && linePtr[1] != '\'') || token == '"') { n = 0; quote = token; while (token == quote) { while (*linePtr != 0 && *linePtr != token) { ch = GetBackslashChar(); if ((ch >= 0) && (instrLen < MAX_BYTSTR)) { bytStr[instrLen++] = ch; n++; } } token = *linePtr; if (token) linePtr++; else Error("Missing close quote"); if (token == quote && *linePtr == quote) // two quotes together { if (instrLen < MAX_BYTSTR) { bytStr[instrLen++] = *linePtr++; n++; } } else token = *linePtr; } // add padding nulls here if ((n & 1) && instrLen < MAX_BYTSTR) bytStr[instrLen++] = 0; } else #endif { linePtr = oldLine; val = Eval(); if ((endian == LITTLE_END) ^ (typ == o_DWRE)) { // little endian if (instrLen < MAX_BYTSTR) bytStr[instrLen++] = val; if (instrLen < MAX_BYTSTR) bytStr[instrLen++] = val >> 8; } else { // big endian if (instrLen < MAX_BYTSTR) bytStr[instrLen++] = val >> 8; if (instrLen < MAX_BYTSTR) bytStr[instrLen++] = val; } } token = GetWord(word); oldLine = linePtr; if (token == ',') { token = GetWord(word); if (token == 0) MissingOperand(); } else if (token) { linePtr = oldLine; Comma(); token = 0; } } if (instrLen >= MAX_BYTSTR) { Error("String too long"); instrLen = MAX_BYTSTR; } instrLen = -instrLen; break; case o_DL: if (endian != LITTLE_END && endian != BIG_END) { Error("CPU endian not defined"); break; } instrLen = 0; oldLine = linePtr; token = GetWord(word); if (token == 0) MissingOperand(); while (token) { #if 1 // enable padded string literals if ((token == '\'' && *linePtr && linePtr[1] != '\'') || token == '"') { n = 0; quote = token; while (token == quote) { while (*linePtr != 0 && *linePtr != token) { ch = GetBackslashChar(); if ((ch >= 0) && (instrLen < MAX_BYTSTR)) { bytStr[instrLen++] = ch; n++; } } token = *linePtr; if (token) linePtr++; else Error("Missing close quote"); if (token == quote && *linePtr == quote) // two quotes together { if (instrLen < MAX_BYTSTR) { bytStr[instrLen++] = *linePtr++; n++; } } else token = *linePtr; } // add padding nulls here while ((n & 3) && instrLen < MAX_BYTSTR) { bytStr[instrLen++] = 0; n++; } } else #endif { linePtr = oldLine; val = Eval(); if ((endian == LITTLE_END) ^ (typ == o_DWRE)) { // little endian if (instrLen < MAX_BYTSTR) bytStr[instrLen++] = val; if (instrLen < MAX_BYTSTR) bytStr[instrLen++] = val >> 8; if (instrLen < MAX_BYTSTR) bytStr[instrLen++] = val >> 16; if (instrLen < MAX_BYTSTR) bytStr[instrLen++] = val >> 24; } else { // big endian if (instrLen < MAX_BYTSTR) bytStr[instrLen++] = val >> 24; if (instrLen < MAX_BYTSTR) bytStr[instrLen++] = val >> 16; if (instrLen < MAX_BYTSTR) bytStr[instrLen++] = val >> 8; if (instrLen < MAX_BYTSTR) bytStr[instrLen++] = val; } } token = GetWord(word); oldLine = linePtr; if (token == ',') { token = GetWord(word); if (token == 0) MissingOperand(); } else if (token) { linePtr = oldLine; Comma(); token = 0; } } if (instrLen >= MAX_BYTSTR) { Error("String too long"); instrLen = MAX_BYTSTR; } instrLen = -instrLen; break; case o_DS: val = Eval(); if (!evalKnown) { Error("Can't use DS pseudo-op with forward-declared length"); break; } oldLine = linePtr; token = GetWord(word); if (token == ',') { if (parm == 1) n = EvalByte(); else n = Eval(); if (val*parm > MAX_BYTSTR) { sprintf(s,"String too long (max %d bytes)",MAX_BYTSTR); Error(s); break; } if (parm == 1) // DS.B for (i=0; i> 24; bytStr[i+1] = n >> 16; bytStr[i+2] = n >> 8; bytStr[i+3] = n; } else { bytStr[i] = n >> 8; bytStr[i+1] = n; } } else { bytStr[i] = n; bytStr[i+1] = n >> 8; if (parm == 4) { bytStr[i+2] = n >> 16; bytStr[i+3] = n >> 24; } } } } instrLen = -val * parm; break; } else if (token) { linePtr = oldLine; Comma(); token = 0; } if (pass == 2) { showAddr = FALSE; // "XXXX (XXXX)" p = ListLoc(locPtr); *p++ = ' '; *p++ = '('; p = ListAddr(p,val); *p++ = ')'; } AddLocPtr(val*parm); break; case o_HEX: instrLen = 0; while (!errFlag && GetWord(word)) { n = strlen(word); for (i=0; i= MAX_BYTSTR) { Error("String too long"); instrLen = MAX_BYTSTR; } instrLen = -instrLen; break; case o_FCC: instrLen = 0; oldLine = linePtr; token = GetWord(word); if (token == 0) MissingOperand(); else if (token == -1) { linePtr = oldLine; val = Eval(); token = GetWord(word); if (val >= MAX_BYTSTR) Error("String too long"); if (!errFlag && (token == ',')) { while (*linePtr >= 0x20 && instrLen < val) bytStr[instrLen++] = *linePtr++; while (instrLen < val) bytStr[instrLen++] = ' '; } else IllegalOperand(); } else { ch = token; while (token) { if (token == ch) { while (token == ch) { while (*linePtr != 0 && *linePtr != token) { if (instrLen < MAX_BYTSTR) bytStr[instrLen++] = *linePtr++; } if (*linePtr) linePtr++; else Error("FCC not terminated properly"); if (*linePtr == token) { if (instrLen < MAX_BYTSTR) bytStr[instrLen++] = *linePtr++; } else token = *linePtr; } } else { linePtr = oldLine; val = EvalByte(); if (instrLen < MAX_BYTSTR) bytStr[instrLen++] = val; } token = GetWord(word); oldLine = linePtr; if (token == ',') { token = GetWord(word); if (token == 0) MissingOperand(); } else if (token) { linePtr = oldLine; Comma(); token = 0; } else if (errFlag) // this is necessary to keep EvalByte() errors token = 0; // from causing phase errors } } if (instrLen >= MAX_BYTSTR) { Error("String too long"); instrLen = MAX_BYTSTR; } instrLen = -instrLen; break; case o_ALIGN_n: val = parm; case o_ALIGN: if (typ == o_ALIGN) val = Eval(); // val must be a power of two if (val <= 0 || val > 65535 || (val & (val - 1)) != 0) { IllegalOperand(); val = 0; } else { i = locPtr & ~(val - 1); if (i != locPtr) val = val - (locPtr - i); // aka val = val + i - locPtr; else val = 0; } if (pass == 2) { showAddr = FALSE; // "XXXX (XXXX)" p = ListLoc(locPtr); *p++ = ' '; *p++ = '('; p = ListAddr(p,val); *p++ = ')'; } AddLocPtr(val); break; case o_END: if (nInclude >= 0) CloseInclude(); else { oldLine = linePtr; if (GetWord(word)) { linePtr = oldLine; val = Eval(); CodeXfer(val); // "XXXX (XXXX)" p = ListLoc(locPtr); *p++ = ' '; *p++ = '('; p = ListAddr(p,val); *p++ = ')'; } sourceEnd = TRUE; } break; case o_Include: GetFName(word); switch(OpenInclude(word)) { case -1: Error("Too many nested INCLUDEs"); break; case 0: sprintf(s,"Unable to open INCLUDE file '%s'",word); Error(s); break; default: break; } break; case o_ENDM: Error("ENDM without MACRO"); break; #ifdef ENABLE_REP case o_REPEND: Error("REPEND without REPEAT"); break; #endif case o_Processor: if (!GetWord(word)) MissingOperand(); else if (!SetCPU(word)) IllegalOperand(); break; default: Error("Unknown opcode"); break; } } void DoLabelOp(int typ, int parm, char *labl) { int val; int i,n; Str255 word,s; int token; Str255 opcode; MacroPtr macro; MacroPtr xmacro; int nparms; SegPtr seg; char *oldLine; // struct MacroRec repList; // repeat text list // MacroLinePtr rep,rep2; // pointer into repeat text list char *p; if (DoCPULabelOp(typ,parm,labl)) return; switch(typ) { case o_EQU: if (labl[0] == 0) Error("Missing label"); else { val = Eval(); // "XXXX (XXXX)" p = listLine; switch(addrWid) { default: case ADDR_16: if (listWid == LIST_24) n=5; else n=4; break; case ADDR_24: n=6; break; case ADDR_32: n=8; break; } for (i = 0; i <= n; i++) *p++ = ' '; *p++ = '='; *p++ = ' '; p = ListAddr(p,val); DefSym(labl,val,parm==1,parm==0); } break; case o_ORG: CodeAbsOrg(Eval()); if (!evalKnown) Warning("Undefined label used in ORG statement"); DefSym(labl,locPtr,FALSE,FALSE); showAddr = TRUE; break; case o_RORG: val = Eval(); CodeRelOrg(val); DefSym(labl,codPtr,FALSE,FALSE); if (pass == 2) { // "XXXX = XXXX" p = ListLoc(codPtr); *p++ = '='; *p++ = ' '; p = ListAddr(p,val); } break; case o_SEG: token = GetWord(word); // get seg name if (token == 0 || token == -1) // must be null or alphanumeric { seg = FindSeg(word); // find segment storage and create if necessary if (!seg) seg = AddSeg(word); // seg -> gen = parm; // copy gen flag from parameter SwitchSeg(seg); DefSym(labl,locPtr,FALSE,FALSE); showAddr = TRUE; } break; case o_SUBR: token = GetWord(word); // get subroutine name strcpy(subrLabl,word); break; case o_REND: if (pass == 2) { // "XXXX = XXXX" p = ListLoc(locPtr); } DefSym(labl,locPtr,FALSE,FALSE); CodeAbsOrg(codPtr); break; case o_LIST: listThisLine = TRUE; // always list this line if (labl[0]) Error("Label not allowed"); GetWord(word); if (strcmp(word,"ON") == 0) listFlag = TRUE; else if (strcmp(word,"OFF") == 0) listFlag = FALSE; else if (strcmp(word,"MACRO") == 0) listMacFlag = TRUE; else if (strcmp(word,"NOMACRO") == 0) listMacFlag = FALSE; else if (strcmp(word,"EXPAND") == 0) expandHexFlag = TRUE; else if (strcmp(word,"NOEXPAND") == 0) expandHexFlag = FALSE; else if (strcmp(word,"SYM") == 0) symtabFlag = TRUE; else if (strcmp(word,"NOSYM") == 0) symtabFlag = FALSE; else if (strcmp(word,"TEMP") == 0) tempSymFlag = TRUE; else if (strcmp(word,"NOTEMP") == 0) tempSymFlag = FALSE; else IllegalOperand(); break; /* case o_PAGE: listThisLine = TRUE; // always list this line if (labl[0]) Error("Label not allowed"); listLineFF = TRUE; break; */ case o_OPT: listThisLine = TRUE; // always list this line if (labl[0]) Error("Label not allowed"); GetWord(word); if (strcmp(word,"LIST") == 0) listFlag = TRUE; else if (strcmp(word,"NOLIST") == 0) listFlag = FALSE; else if (strcmp(word,"MACRO") == 0) listMacFlag = TRUE; else if (strcmp(word,"NOMACRO") == 0) listMacFlag = FALSE; else if (strcmp(word,"EXPAND") == 0) expandHexFlag = TRUE; else if (strcmp(word,"NOEXPAND") == 0) expandHexFlag = FALSE; else if (strcmp(word,"SYM") == 0) symtabFlag = TRUE; else if (strcmp(word,"NOSYM") == 0) symtabFlag = FALSE; else if (strcmp(word,"TEMP") == 0) tempSymFlag = TRUE; else if (strcmp(word,"NOTEMP") == 0) tempSymFlag = FALSE; else Error("Illegal option"); break; case o_ERROR: if (labl[0]) Error("Label not allowed"); while (*linePtr == ' ' || *linePtr == '\t') linePtr++; Error(linePtr); break; case o_MACRO: // see if label already provided if (labl[0] == 0) { // get next word on line for macro name if (GetWord(labl) != -1) { Error("Macro name requried"); break; } // optional comma after macro name oldLine = linePtr; if (GetWord(word) != ',') linePtr = oldLine; } // don't allow temporary labels as macro names if (strchr(labl,'@') || strchr(labl,'.')) { Error("Can not use temporary labels as macro names"); break; } macro = FindMacro(labl); if (macro && macro -> def) Error("Macro multiply defined"); else { if (macro == NULL) { macro = AddMacro(labl); nparms = 0; token = GetWord(word); while (token == -1) { nparms++; if (nparms > MAXMACPARMS) { Error("Too many macro parameters"); macro -> toomany = TRUE; break; } AddMacroParm(macro,word); token = GetWord(word); if (token == ',') token = GetWord(word); } if (word[0] && !errFlag) Error("Illegal operand"); } if (pass == 2) { macro -> def = TRUE; if (macro -> toomany) Error("Too many macro parameters"); } macroCondLevel = 0; i = ReadSourceLine(line, sizeof(line)); while (i && typ != o_ENDM) { if ((pass == 2 || cl_ListP1) && (listFlag || errFlag)) ListOut(TRUE); CopyListLine(); // skip initial formfeeds linePtr = line; while (*linePtr == 12) linePtr++; // get label labl[0] = 0; if (isalphanum(*linePtr) || *linePtr == '@' || *linePtr == '.') { token = GetWord(labl); if (token) showAddr = TRUE; while (*linePtr == ' ' || *linePtr == '\t') linePtr++; if (labl[0]) { if (token == '@' || token == '.') { GetWord(word); if (token == '.' && subrLabl[0]) strcpy(labl,subrLabl); else strcpy(labl,lastLabl); labl[strlen(labl)+1] = 0; labl[strlen(labl)] = token; strcat(labl,word); // labl = lastLabl + "@" + word; } else strcpy(lastLabl,labl); } if (*linePtr == ':' && linePtr[1] != '=') linePtr++; } typ = 0; GetFindOpcode(opcode, &typ, &parm, &xmacro); switch(typ) { case o_IF: if (pass == 1) AddMacroLine(macro,line); macroCondLevel++; break; case o_ENDIF: if (pass == 1) AddMacroLine(macro,line); if (macroCondLevel) macroCondLevel--; else Error("ENDIF without IF in macro definition"); break; case o_END: Error("END not allowed inside a macro"); break; case o_ENDM: if (pass == 1 && labl[0]) AddMacroLine(macro,labl); break; default: if (pass == 1) AddMacroLine(macro,line); break; } if (typ != o_ENDM) i = ReadSourceLine(line, sizeof(line)); } if (macroCondLevel) Error("IF block without ENDIF in macro definition"); if (typ != o_ENDM) Error("Missing ENDM"); } break; case o_IF: if (labl[0]) Error("Label not allowed"); if (condLevel >= MAX_COND) Error("IF statements nested too deeply"); else { condLevel++; condState[condLevel] = 0; // this block false but level not permanently failed val = Eval(); if (!errFlag && val != 0) condState[condLevel] = condTRUE; // this block true } break; case o_ELSE: // previous IF was true, so this section stays off if (labl[0]) Error("Label not allowed"); if (condLevel == 0) Error("ELSE outside of IF block"); else if (condLevel < MAX_COND && (condState[condLevel] & condELSE)) Error("Multiple ELSE statements in an IF block"); else { condState[condLevel] = condELSE | condFAIL; // ELSE encountered, permanent fail // condState[condLevel] |= condELSE; // ELSE encountered // condState[condLevel] |= condFAIL; // this level permanently failed // condState[condLevel] &= ~condTRUE; // this block false } break; case o_ELSIF: // previous IF was true, so this section stays off if (labl[0]) Error("Label not allowed"); if (condLevel == 0) Error("ELSIF outside of IF block"); else if (condLevel < MAX_COND && (condState[condLevel] & condELSE)) Error("Multiple ELSE statements in an IF block"); else { // i = Eval(); // evaluate condition and ignore result EatIt(); // just ignore the conditional expression condState[condLevel] |= condFAIL; // this level permanently failed condState[condLevel] &= ~condTRUE; // this block false } break; case o_ENDIF: // previous ELSE was true, now exiting it if (labl[0]) Error("Label not allowed"); if (condLevel == 0) Error("ENDIF outside of IF block"); else condLevel--; break; #ifdef ENABLE_REP // still under construction // notes: // REPEAT pseudo-op should act like an inline macro // 1) collect all lines into new macro level // 2) copy parameters from previous macro level (if any) // 3) set repeat count for this macro level (repeat count is set to 0 for plain macro) // 4) when macro ends, decrement repeat count and start over if count > 0 // 5) don't forget to dispose of the temp macro and its lines when done! case o_REPEAT: if (labl[0]) { DefSym(labl,locPtr,FALSE,FALSE); showAddr = TRUE; } // get repeat count val = Eval(); if (!errFlag) if (val < 0) IllegalOperand(); if (!errFlag) { repList -> text = NULL; // *** read line // *** while line not REPEND // *** add line to repeat buffer macroCondLevel = 0; i = ReadSourceLine(line, sizeof(line)); while (i && typ != o_REPEND) { if ((pass == 2 || cl_ListP1) && (listFlag || errFlag)) ListOut(TRUE); CopyListLine(); // skip initial formfeeds linePtr = line; while (*linePtr == 12) linePtr++; // get label labl[0] = 0; if (isalphanum(*linePtr) || *linePtr == '@' || *linePtr == '.') { token = GetWord(labl); if (token) showAddr = TRUE; while (*linePtr == ' ' || *linePtr == '\t') linePtr++; if (labl[0]) { if (token == '@' || token == '.') { GetWord(word); if (token == '.' && subrLabl[0]) strcpy(labl,subrLabl); else strcpy(labl,lastLabl); labl[strlen(labl)+1] = 0; labl[strlen(labl)] = token; strcat(labl,word); // labl = lastLabl + "@" + word; } else strcpy(lastLabl,labl); } if (*linePtr == ':' && linePtr[1] != '=') linePtr++; } typ = 0; GetFindOpcode(opcode, &typ, &parm, &xmacro); switch(typ) { case o_IF: if (pass == 1) AddMacroLine(&replist,line); macroCondLevel++; break; case o_ENDIF: if (pass == 1) AddMacroLine(&replist,line); if (macroCondLevel) macroCondLevel--; else Error("ENDIF without IF in REPEAT block"); break; case o_END: Error("END not allowed inside REPEAT block"); break; case o_ENDM: if (pass == 1 && labl[0]) AddMacroLine(&replist,labl); break; default: if (pass == 1) AddMacroLine(&replist,line); break; } if (typ != o_ENDM) i = ReadSourceLine(line, sizeof(line)); } if (macroCondLevel) Error("IF block without ENDIF in REPEAT block"); if (typ != o_REPEND) Error("Missing REPEND"); if (!errFlag) { // *** while (val--) // *** for each line // *** doline() } // free repeat line buffer while (replist) { rep = replist->next; free(replist); replist = rep; } } break; #endif case o_Incbin: DefSym(labl,locPtr,FALSE,FALSE); GetFName(word); val = 0; // open binary file incbin = fopen(word, "r"); if (incbin) { // while not EOF do { // n = count of read up to MAX_BYTSTR bytes into bytStr n = fread(bytStr, 1, MAX_BYTSTR, incbin); if (n>0) { // write data out to the object file for (i=0; i0); if (n<0) sprintf(s,"Error reading INCBIN file '%s'",word); if (pass == 2) { // "XXXX (XXXX)" p = ListLoc(locPtr-val); *p++ = ' '; *p++ = '('; p = ListAddr(p,val); *p++ = ')'; } } else { sprintf(s,"Unable to open INCBIN file '%s'",word); Error(s); } // close binary file if (incbin) fclose(incbin); incbin = NULL; break; case o_WORDSIZE: if (labl[0]) Error("Label not allowed"); val = Eval(); if (evalKnown) { Error("Forward reference not allowed in WORDSIZE"); } else if (!errFlag) { if (val == 0) SetWordSize(wordSize); else if (val < 1 || val > 64) Error("WORDSIZE must be in the range of 0..64"); else SetWordSize(val); } break; default: Error("Unknown opcode"); break; } } void DoLine() { Str255 labl; Str255 opcode; int typ; int parm; int i; Str255 word; int token; MacroPtr macro; char *oldLine; char *p; int numhex; bool firstLine; errFlag = FALSE; warnFlag = FALSE; instrLen = 0; showAddr = FALSE; listThisLine = listFlag; firstLine = TRUE; CopyListLine(); // skip initial formfeeds linePtr = line; while (*linePtr == 12) linePtr++; // look for label at beginning of line labl[0] = 0; if (isalphaul(*linePtr) || *linePtr == '$' || *linePtr == '@' || *linePtr == '.') { token = GetWord(labl); oldLine = linePtr; if (token) showAddr = TRUE; while (*linePtr == ' ' || *linePtr == '\t') linePtr++; if (labl[0]) { if (token == '@' || token == '.') { GetWord(word); if (token == '.' && FindOpcodeTab((OpcdPtr) &opcdTab2, word, &typ, &parm) ) linePtr = oldLine; else if (token == '.' && FindCPU(word)) linePtr = line; else { if (token == '.' && subrLabl[0]) strcpy(labl,subrLabl); else strcpy(labl,lastLabl); labl[strlen(labl)+1] = 0; labl[strlen(labl)] = token; strcat(labl,word); // labl = lastLabl + "@" + word; } } else strcpy(lastLabl,labl); } if (*linePtr == ':' && linePtr[1] != '=') linePtr++; } if (!(condState[condLevel] & condTRUE)) { listThisLine = FALSE; // inside failed IF statement if (GetFindOpcode(opcode, &typ, &parm, ¯o)) { switch(typ) { case o_IF: // nested IF inside failed IF should stay failed if (condState[condLevel-1] & condTRUE) listThisLine = listFlag; if (condLevel >= MAX_COND) Error("IF statements nested too deeply"); else { condLevel++; condState[condLevel] = condFAIL; // this level false and permanently failed } break; case o_ELSE: if (condState[condLevel-1] & condTRUE) listThisLine = listFlag; if (!(condState[condLevel] & (condTRUE | condFAIL))) { // previous IF was false listThisLine = listFlag; if (condLevel < MAX_COND && (condState[condLevel] & condELSE)) Error("Multiple ELSE statements in an IF block"); else condState[condLevel] |= condTRUE; } condState[condLevel] |= condELSE; break; case o_ELSIF: if (condState[condLevel-1] & condTRUE) listThisLine = listFlag; if (!(condState[condLevel] & (condTRUE | condFAIL))) { // previous IF was false listThisLine = listFlag; if (condLevel < MAX_COND && (condState[condLevel] & condELSE)) Error("Multiple ELSE statements in an IF block"); else { i = Eval(); if (!errFlag && i != 0) condState[condLevel] |= condTRUE; } } break; case o_ENDIF: if (condLevel == 0) Error("ENDIF outside of IF block"); else { condLevel--; if (condState[condLevel] & condTRUE) listThisLine = listFlag; } break; default: // ignore any other lines break; } } if ((pass == 2 || cl_ListP1) && listThisLine && (errFlag || listMacFlag || !macLineFlag)) ListOut(TRUE); } else { if (!GetFindOpcode(opcode, &typ, &parm, ¯o) && !opcode[0]) { // line with label only DefSym(labl,locPtr / wordDiv,FALSE,FALSE); } else { if (typ == o_Illegal) { if (opcode[0] == '.' && SetCPU(opcode+1)) /* successfully changed CPU type */; else { sprintf(word,"Illegal opcode '%s'",opcode); Error(word); } } else if (typ == o_MacName) { if (macPtr[macLevel] && macLevel >= MAX_MACRO) Error("Macros nested too deeply"); #if 1 else if (pass == 2 && !macro -> def) Error("Macro has not been defined yet"); #endif else { if (macPtr[macLevel]) macLevel++; macPtr [macLevel] = macro; macLine[macLevel] = macro -> text; #ifdef ENABLE_REP macRepeat[macLevel] = 0; #endif GetMacParms(macro); showAddr = TRUE; DefSym(labl,locPtr,FALSE,FALSE); } } else if (typ >= o_LabelOp) { showAddr = FALSE; DoLabelOp(typ,parm,labl); } else { showAddr = TRUE; DefSym(labl,locPtr,FALSE,FALSE); DoOpcode(typ, parm); } if (typ != o_Illegal && typ != o_MacName) if (!errFlag && GetWord(word)) Error("Too many operands"); } if (pass == 1 && !cl_ListP1) AddLocPtr(abs(instrLen)); else { p = listLine; if (showAddr) p = ListLoc(locPtr); else switch(addrWid) { default: case ADDR_16: p = listLine + 5; break; case ADDR_24: p = listLine + 7; break; case ADDR_32: p = listLine + 9; break; } // determine width of hex data area if (listWid == LIST_16) numhex = 5; else // listWid == LIST_24 { switch(addrWid) { default: case ADDR_16: case ADDR_24: numhex = 8; break; case ADDR_32: numhex = 6; break; } } if (instrLen>0) // positive instrLen for CPU instruction formatting { // determine start of hex data area switch(addrWid) { default: case ADDR_16: if (listWid == LIST_24) p = listLine + 6; else p = listLine + 5; break; case ADDR_24: p = listLine + 7; break; case ADDR_32: p = listLine + 9; break; } // special case because 24-bit address usually can't fit // 8 bytes of code with operand spacing if (addrWid == ADDR_24 && listWid == LIST_24) numhex = 6; if (hexSpaces & 1) { *p++ = ' '; } for (i = 0; i < instrLen; i++) { if (listWid == LIST_24) if (hexSpaces & (1< 0 && i % numhex == 0) { if (listThisLine && (errFlag || listMacFlag || !macLineFlag)) { ListOut(firstLine); firstLine = FALSE; } if (listWid == LIST_24) strcpy(listLine, " "); // 24 blanks else strcpy(listLine, " "); // 16 blanks p = ListLoc(locPtr); } p = ListByte(p,bytStr[i]); } CodeOut(bytStr[i]); } } else if (instrLen<0) // negative instrLen for generic data formatting { instrLen = abs(instrLen); for (i = 0; i < instrLen; i++) { if (i 0 && i % numhex == 0) { if (listThisLine && (errFlag || listMacFlag || !macLineFlag)) { ListOut(firstLine); firstLine = FALSE; } if (listWid == LIST_24) strcpy(listLine, " "); // 24 blanks else strcpy(listLine, " "); // 16 blanks p = ListLoc(locPtr); } if (numhex == 6 && (i%numhex) == 2) *p++ = ' '; if (numhex == 6 && (i%numhex) == 4) *p++ = ' '; if (numhex == 8 && (i%numhex) == 4 && addrWid != ADDR_24) *p++ = ' '; p = ListByte(p,bytStr[i]); if (i>=numhex) *p = 0; } CodeOut(bytStr[i]); } } if (listThisLine && (errFlag || listMacFlag || !macLineFlag)) ListOut(firstLine); } } } void DoPass() { Str255 opcode; int typ; int parm; int i; MacroPtr macro; SegPtr seg; fseek(source, 0, SEEK_SET); // rewind source file sourceEnd = FALSE; lastLabl[0] = 0; subrLabl[0] = 0; fprintf(stderr,"Pass %d\n",pass); if (cl_ListP1) fprintf(listing,"Pass %d\n",pass); errCount = 0; condLevel = 0; condState[condLevel] = condTRUE; // top level always true listFlag = TRUE; listMacFlag = FALSE; expandHexFlag = TRUE; symtabFlag = TRUE; tempSymFlag = TRUE; linenum = 0; macLevel = 0; macUniqueID = 0; macCurrentID[0] = 0; curAsm = NULL; endian = UNKNOWN_END; opcdTab = NULL; listWid = LIST_24; addrWid = ADDR_32; wordSize = 8; SetWordSize(wordSize); SetCPU(defCPU); addrMax = addrWid; // reset all code pointers CodeAbsOrg(0); seg = segTab; while (seg) { seg -> cod = 0; seg -> loc = 0; seg = seg -> next; } curSeg = nullSeg; PassInit(); i = ReadSourceLine(line, sizeof(line)); while (i && !sourceEnd) { DoLine(); i = ReadSourceLine(line, sizeof(line)); } if (condLevel != 0) Error("IF block without ENDIF"); if (pass == 2) CodeEnd(); // Put the lines after the END statement into the listing file // while still checking for listing control statements. Ignore // any lines which have invalid syntax, etc., because whatever // is found after an END statement should esentially be ignored. if (pass == 2 || cl_ListP1) { while (i) { listThisLine = listFlag; CopyListLine(); if (line[0]==' ' || line[0]=='\t') // ignore labels (this isn't the right way) { GetFindOpcode(opcode, &typ, &parm, ¯o); if (typ == o_LIST || typ == o_OPT) { DoLabelOp(typ,parm,""); } } if (listThisLine) ListOut(TRUE); i = ReadSourceLine(line, sizeof(line)); } } } // -------------------------------------------------------------- // initialization and parameters void stdversion(void) { fprintf(stderr,"%s version %s\n",VERSION_NAME,VERSION); fprintf(stderr,"%s\n",COPYRIGHT); } void usage(void) { stdversion(); fprintf(stderr,"\n"); fprintf(stderr,"Usage:\n"); fprintf(stderr," %s [options] srcfile\n",progname); fprintf(stderr,"\n"); fprintf(stderr,"Options:\n"); fprintf(stderr," -- end of options\n"); fprintf(stderr," -e show errors to screen\n"); fprintf(stderr," -w show warnings to screen\n"); // fprintf(stderr," -1 output listing during first pass\n"); fprintf(stderr," -l [filename] make a listing file, default is srcfile.lst\n"); fprintf(stderr," -o [filename] make an object file, default is srcfile.hex or srcfile.s9\n"); fprintf(stderr," -d label[[:]=value] define a label, and assign an optional value\n"); // fprintf(stderr," -9 output object file in Motorola S9 format (16-bit address)\n"); fprintf(stderr," -s9 output object file in Motorola S9 format (16-bit address)\n"); fprintf(stderr," -s19 output object file in Motorola S9 format (16-bit address)\n"); fprintf(stderr," -s28 output object file in Motorola S9 format (24-bit address)\n"); fprintf(stderr," -s37 output object file in Motorola S9 format (32-bit address)\n"); fprintf(stderr," -b [baseaddr] output object file as binary with optional base address\n"); fprintf(stderr," -c send object code to stdout\n"); fprintf(stderr," -C cputype specify default CPU type (currently "); if (defCPU[0]) fprintf(stderr,"%s",defCPU); else fprintf(stderr,"no default"); fprintf(stderr,")\n"); exit(1); } void getopts(int argc, char * const argv[]) { int ch; int val; Str255 labl,word; bool setSym; int token; int neg; while ((ch = getopt(argc, argv, "ew19b:cd:l:o:s:C:?")) != -1) { switch (ch) { case 'e': cl_Err = TRUE; break; case 'w': cl_Warn = TRUE; break; case '1': cl_ListP1 = TRUE; break; case '9': // -9 option is deprecated cl_S9type = 9; cl_ObjType = OBJ_S9; break; case 's': if (optarg[0] == '9' && optarg[1] == 0) cl_S9type = 9; else if (optarg[0] == '1' && optarg[1] == '9' && optarg[2] == 0) cl_S9type = 19; else if (optarg[0] == '2' && optarg[1] == '8' && optarg[2] == 0) cl_S9type = 28; else if (optarg[0] == '3' && optarg[1] == '7' && optarg[2] == 0) cl_S9type = 37; else usage(); cl_ObjType = OBJ_S9; break; case 'b': cl_ObjType = OBJ_BIN; cl_Binbase = 0; if (optarg[0] =='-') { optarg = ""; optind--; } else if (*optarg) cl_Binbase = EvalNum(optarg); break; case 'c': if (cl_Obj) { fprintf(stderr,"%s: Conflicting options: -c can not be used with -o\n",progname); usage(); } cl_Stdout = TRUE; break; case 'd': strncpy(line, optarg, 255); linePtr = line; GetWord(labl); val = 0; setSym = FALSE; token = GetWord(word); if (token == ':') { setSym = TRUE; token = GetWord(word); } if (token == '=') { neg = 1; if (GetWord(word) == '-') { neg = -1; GetWord(word); } val = neg * EvalNum(word); } DefSym(labl,val,setSym,!setSym); break; case 'l': cl_List = TRUE; if (optarg[0] =='-') { optarg = ""; optind--; } strncpy(cl_ListName, optarg, 255); break; case 'o': if (cl_Stdout) { fprintf(stderr,"%s: Conflicting options: -o can not be used with -c\n",progname); usage(); } cl_Obj = TRUE; if (optarg[0] =='-') { optarg = ""; optind--; } strncpy(cl_ObjName, optarg, 255); break; case 'C': strncpy(word, optarg, 255); Uprcase(word); if (!FindCPU(word)) { fprintf(stderr,"CPU type '%s' unknown\n",word); usage(); } strcpy(defCPU, word); break; case '?': default: usage(); } } argc -= optind; argv += optind; if (cl_Stdout && cl_ObjType == OBJ_BIN) { fprintf(stderr,"%s: Conflicting options: -b can not be used with -c\n",progname); usage(); } #if 1 // -b or -9 must force -o! if (cl_ObjType != OBJ_HEX && !cl_Stdout && !cl_Obj) cl_Obj = TRUE; #endif // now argc is the number of remaining arguments // and argv[0] is the first remaining argument if (argc != 1) usage(); strncpy(cl_SrcName, argv[0], 255); // note: this won't work if there's a single-char filename in the current directory! if (cl_SrcName[0] == '?' && cl_SrcName[1] == 0) usage(); if (cl_List && cl_ListName[0] == 0) { strncpy(cl_ListName, cl_SrcName, 255-4); strcat (cl_ListName, ".lst"); } if (cl_Obj && cl_ObjName [0] == 0) { switch(cl_ObjType) { case OBJ_S9: strncpy(cl_ObjName, cl_SrcName, 255-3); sprintf(word,".s%d",cl_S9type); strcat (cl_ObjName, word); break; case OBJ_BIN: strncpy(cl_ObjName, cl_SrcName, 255-4); strcat (cl_ObjName, ".bin"); break; default: case OBJ_HEX: strncpy(cl_ObjName, cl_SrcName, 255-4); strcat (cl_ObjName, ".hex"); break; } } } int main(int argc, char * const argv[]) { int i; // initialize and get parms progname = argv[0]; pass = 0; symTab = NULL; xferAddr = 0; xferFound = FALSE; macroTab = NULL; macPtr[0] = NULL; macLine[0] = NULL; segTab = NULL; nullSeg = AddSeg(""); curSeg = nullSeg; cl_Err = FALSE; cl_Warn = FALSE; cl_List = FALSE; cl_Obj = FALSE; cl_ObjType = OBJ_HEX; cl_ListP1 = FALSE; asmTab = NULL; cpuTab = NULL; defCPU[0] = 0; nInclude = -1; for (i=0; i