Émuler un processeur MOS 6502

Je pensais que jirais de lavant et publierais ma propre implémentation. Cest COMPLÈTEMENT sans jeu, mais cest une implémentation complète.

• 668 lignes de C. (sans compter les lignes vides ou les lignes avec seulement des commentaires)
• Prend en charge (je pense ) toutes les instructions non documentées.
• Prend en charge BCD.
• synchronisation du cycle dhorloge du processeur. (y compris les ajustements sur certains wraps de limites de page)
• Peut exécuter des instructions en une seule étape ou en spécifiant le nombre de graduations.
• Prend en charge laccrochage dune fonction externe à appeler après chaque instruction est exectué. Cétait parce que cétait à lorigine pour un émulateur NES et que je lai utilisé pour la synchronisation audio.

/* Fake6502 CPU emulator core v1.1 ******************* * (c)2011-2013 Mike Chambers * *****************************************************/ #include <stdio.h> #include <stdint.h> //externally supplied functions extern uint8_t read6502(uint16_t address); extern void write6502(uint16_t address, uint8_t value); //6502 defines #define UNDOCUMENTED //when this is defined, undocumented opcodes are handled. //otherwise, they"re simply treated as NOPs. //#define NES_CPU //when this is defined, the binary-coded decimal (BCD) //status flag is not honored by ADC and SBC. the 2A03 //CPU in the Nintendo Entertainment System does not //support BCD operation. #define FLAG_CARRY 0x01 #define FLAG_ZERO 0x02 #define FLAG_INTERRUPT 0x04 #define FLAG_DECIMAL 0x08 #define FLAG_BREAK 0x10 #define FLAG_CONSTANT 0x20 #define FLAG_OVERFLOW 0x40 #define FLAG_SIGN 0x80 #define BASE_STACK 0x100 #define saveaccum(n) a = (uint8_t)((n) & 0x00FF) //flag modifier macros #define setcarry() status |= FLAG_CARRY #define clearcarry() status &= (~FLAG_CARRY) #define setzero() status |= FLAG_ZERO #define clearzero() status &= (~FLAG_ZERO) #define setinterrupt() status |= FLAG_INTERRUPT #define clearinterrupt() status &= (~FLAG_INTERRUPT) #define setdecimal() status |= FLAG_DECIMAL #define cleardecimal() status &= (~FLAG_DECIMAL) #define setoverflow() status |= FLAG_OVERFLOW #define clearoverflow() status &= (~FLAG_OVERFLOW) #define setsign() status |= FLAG_SIGN #define clearsign() status &= (~FLAG_SIGN) //flag calculation macros #define zerocalc(n) {\ if ((n) & 0x00FF) clearzero();\ else setzero();\ } #define signcalc(n) {\ if ((n) & 0x0080) setsign();\ else clearsign();\ } #define carrycalc(n) {\ if ((n) & 0xFF00) setcarry();\ else clearcarry();\ } #define overflowcalc(n, m, o) { /* n = result, m = accumulator, o = memory */ \ if (((n) ^ (uint16_t)(m)) & ((n) ^ (o)) & 0x0080) setoverflow();\ else clearoverflow();\ } //6502 CPU registers uint16_t pc; uint8_t sp, a, x, y, status = FLAG_CONSTANT; //helper variables uint64_t instructions = 0; //keep track of total instructions executed uint32_t clockticks6502 = 0, clockgoal6502 = 0; uint16_t oldpc, ea, reladdr, value, result; uint8_t opcode, oldstatus; //a few general functions used by various other functions void push16(uint16_t pushval) { write6502(BASE_STACK + sp, (pushval >> 8) & 0xFF); write6502(BASE_STACK + ((sp - 1) & 0xFF), pushval & 0xFF); sp -= 2; } void push8(uint8_t pushval) { write6502(BASE_STACK + sp--, pushval); } uint16_t pull16() { uint16_t temp16; temp16 = read6502(BASE_STACK + ((sp + 1) & 0xFF)) | ((uint16_t)read6502(BASE_STACK + ((sp + 2) & 0xFF)) << 8); sp += 2; return(temp16); } uint8_t pull8() { return (read6502(BASE_STACK + ++sp)); } void reset6502() { pc = (uint16_t)read6502(0xFFFC) | ((uint16_t)read6502(0xFFFD) << 8); a = 0; x = 0; y = 0; sp = 0xFD; status |= FLAG_CONSTANT; } static void (*addrtable[256])(); static void (*optable[256])(); uint8_t penaltyop, penaltyaddr; //addressing mode functions, calculates effective addresses static void imp() { //implied } static void acc() { //accumulator } static void imm() { //immediate ea = pc++; } static void zp() { //zero-page ea = (uint16_t)read6502((uint16_t)pc++); } static void zpx() { //zero-page,X ea = ((uint16_t)read6502((uint16_t)pc++) + (uint16_t)x) & 0xFF; //zero-page wraparound } static void zpy() { //zero-page,Y ea = ((uint16_t)read6502((uint16_t)pc++) + (uint16_t)y) & 0xFF; //zero-page wraparound } static void rel() { //relative for branch ops (8-bit immediate value, sign-extended) reladdr = (uint16_t)read6502(pc++); if (reladdr & 0x80) reladdr |= 0xFF00; } static void abso() { //absolute ea = (uint16_t)read6502(pc) | ((uint16_t)read6502(pc+1) << 8); pc += 2; } static void absx() { //absolute,X uint16_t startpage; ea = ((uint16_t)read6502(pc) | ((uint16_t)read6502(pc+1) << 8)); startpage = ea & 0xFF00; ea += (uint16_t)x; if (startpage != (ea & 0xFF00)) { //one cycle penlty for page-crossing on some opcodes penaltyaddr = 1; } pc += 2; } static void absy() { //absolute,Y uint16_t startpage; ea = ((uint16_t)read6502(pc) | ((uint16_t)read6502(pc+1) << 8)); startpage = ea & 0xFF00; ea += (uint16_t)y; if (startpage != (ea & 0xFF00)) { //one cycle penlty for page-crossing on some opcodes penaltyaddr = 1; } pc += 2; } static void ind() { //indirect uint16_t eahelp, eahelp2; eahelp = (uint16_t)read6502(pc) | (uint16_t)((uint16_t)read6502(pc+1) << 8); eahelp2 = (eahelp & 0xFF00) | ((eahelp + 1) & 0x00FF); //replicate 6502 page-boundary wraparound bug ea = (uint16_t)read6502(eahelp) | ((uint16_t)read6502(eahelp2) << 8); pc += 2; } static void indx() { // (indirect,X) uint16_t eahelp; eahelp = (uint16_t)(((uint16_t)read6502(pc++) + (uint16_t)x) & 0xFF); //zero-page wraparound for table pointer ea = (uint16_t)read6502(eahelp & 0x00FF) | ((uint16_t)read6502((eahelp+1) & 0x00FF) << 8); } static void indy() { // (indirect),Y uint16_t eahelp, eahelp2, startpage; eahelp = (uint16_t)read6502(pc++); eahelp2 = (eahelp & 0xFF00) | ((eahelp + 1) & 0x00FF); //zero-page wraparound ea = (uint16_t)read6502(eahelp) | ((uint16_t)read6502(eahelp2) << 8); startpage = ea & 0xFF00; ea += (uint16_t)y; if (startpage != (ea & 0xFF00)) { //one cycle penlty for page-crossing on some opcodes penaltyaddr = 1; } } static uint16_t getvalue() { if (addrtable[opcode] == acc) return((uint16_t)a); else return((uint16_t)read6502(ea)); } static void putvalue(uint16_t saveval) { if (addrtable[opcode] == acc) a = (uint8_t)(saveval & 0x00FF); else write6502(ea, (saveval & 0x00FF)); } //instruction handler functions static void adc() { penaltyop = 1; value = getvalue(); result = (uint16_t)a + value + (uint16_t)(status & FLAG_CARRY); carrycalc(result); zerocalc(result); overflowcalc(result, a, value); signcalc(result); #ifndef NES_CPU if (status & FLAG_DECIMAL) { clearcarry(); if ((a & 0x0F) > 0x09) { a += 0x06; } if ((a & 0xF0) > 0x90) { a += 0x60; setcarry(); } clockticks6502++; } #endif saveaccum(result); } static void and() { penaltyop = 1; value = getvalue(); result = (uint16_t)a & value; zerocalc(result); signcalc(result); saveaccum(result); } static void asl() { value = getvalue(); result = value << 1; carrycalc(result); zerocalc(result); signcalc(result); putvalue(result); } static void bcc() { if ((status & FLAG_CARRY) == 0) { oldpc = pc; pc += reladdr; if ((oldpc & 0xFF00) != (pc & 0xFF00)) clockticks6502 += 2; //check if jump crossed a page boundary else clockticks6502++; } } static void bcs() { if ((status & FLAG_CARRY) == FLAG_CARRY) { oldpc = pc; pc += reladdr; if ((oldpc & 0xFF00) != (pc & 0xFF00)) clockticks6502 += 2; //check if jump crossed a page boundary else clockticks6502++; } } static void beq() { if ((status & FLAG_ZERO) == FLAG_ZERO) { oldpc = pc; pc += reladdr; if ((oldpc & 0xFF00) != (pc & 0xFF00)) clockticks6502 += 2; //check if jump crossed a page boundary else clockticks6502++; } } static void bit() { value = getvalue(); result = (uint16_t)a & value; zerocalc(result); status = (status & 0x3F) | (uint8_t)(value & 0xC0); } static void bmi() { if ((status & FLAG_SIGN) == FLAG_SIGN) { oldpc = pc; pc += reladdr; if ((oldpc & 0xFF00) != (pc & 0xFF00)) clockticks6502 += 2; //check if jump crossed a page boundary else clockticks6502++; } } static void bne() { if ((status & FLAG_ZERO) == 0) { oldpc = pc; pc += reladdr; if ((oldpc & 0xFF00) != (pc & 0xFF00)) clockticks6502 += 2; //check if jump crossed a page boundary else clockticks6502++; } } static void bpl() { if ((status & FLAG_SIGN) == 0) { oldpc = pc; pc += reladdr; if ((oldpc & 0xFF00) != (pc & 0xFF00)) clockticks6502 += 2; //check if jump crossed a page boundary else clockticks6502++; } } static void brk() { pc++; push16(pc); //push next instruction address onto stack push8(status | FLAG_BREAK); //push CPU status to stack setinterrupt(); //set interrupt flag pc = (uint16_t)read6502(0xFFFE) | ((uint16_t)read6502(0xFFFF) << 8); } static void bvc() { if ((status & FLAG_OVERFLOW) == 0) { oldpc = pc; pc += reladdr; if ((oldpc & 0xFF00) != (pc & 0xFF00)) clockticks6502 += 2; //check if jump crossed a page boundary else clockticks6502++; } } static void bvs() { if ((status & FLAG_OVERFLOW) == FLAG_OVERFLOW) { oldpc = pc; pc += reladdr; if ((oldpc & 0xFF00) != (pc & 0xFF00)) clockticks6502 += 2; //check if jump crossed a page boundary else clockticks6502++; } } static void clc() { clearcarry(); } static void cld() { cleardecimal(); } static void cli() { clearinterrupt(); } static void clv() { clearoverflow(); } static void cmp() { penaltyop = 1; value = getvalue(); result = (uint16_t)a - value; if (a >= (uint8_t)(value & 0x00FF)) setcarry(); else clearcarry(); if (a == (uint8_t)(value & 0x00FF)) setzero(); else clearzero(); signcalc(result); } static void cpx() { value = getvalue(); result = (uint16_t)x - value; if (x >= (uint8_t)(value & 0x00FF)) setcarry(); else clearcarry(); if (x == (uint8_t)(value & 0x00FF)) setzero(); else clearzero(); signcalc(result); } static void cpy() { value = getvalue(); result = (uint16_t)y - value; if (y >= (uint8_t)(value & 0x00FF)) setcarry(); else clearcarry(); if (y == (uint8_t)(value & 0x00FF)) setzero(); else clearzero(); signcalc(result); } static void dec() { value = getvalue(); result = value - 1; zerocalc(result); signcalc(result); putvalue(result); } static void dex() { x--; zerocalc(x); signcalc(x); } static void dey() { y--; zerocalc(y); signcalc(y); } static void eor() { penaltyop = 1; value = getvalue(); result = (uint16_t)a ^ value; zerocalc(result); signcalc(result); saveaccum(result); } static void inc() { value = getvalue(); result = value + 1; zerocalc(result); signcalc(result); putvalue(result); } static void inx() { x++; zerocalc(x); signcalc(x); } static void iny() { y++; zerocalc(y); signcalc(y); } static void jmp() { pc = ea; } static void jsr() { push16(pc - 1); pc = ea; } static void lda() { penaltyop = 1; value = getvalue(); a = (uint8_t)(value & 0x00FF); zerocalc(a); signcalc(a); } static void ldx() { penaltyop = 1; value = getvalue(); x = (uint8_t)(value & 0x00FF); zerocalc(x); signcalc(x); } static void ldy() { penaltyop = 1; value = getvalue(); y = (uint8_t)(value & 0x00FF); zerocalc(y); signcalc(y); } static void lsr() { value = getvalue(); result = value >> 1; if (value & 1) setcarry(); else clearcarry(); zerocalc(result); signcalc(result); putvalue(result); } static void nop() { switch (opcode) { case 0x1C: case 0x3C: case 0x5C: case 0x7C: case 0xDC: case 0xFC: penaltyop = 1; break; } } static void ora() { penaltyop = 1; value = getvalue(); result = (uint16_t)a | value; zerocalc(result); signcalc(result); saveaccum(result); } static void pha() { push8(a); } static void php() { push8(status | FLAG_BREAK); } static void pla() { a = pull8(); zerocalc(a); signcalc(a); } static void plp() { status = pull8() | FLAG_CONSTANT; } static void rol() { value = getvalue(); result = (value << 1) | (status & FLAG_CARRY); carrycalc(result); zerocalc(result); signcalc(result); putvalue(result); } static void ror() { value = getvalue(); result = (value >> 1) | ((status & FLAG_CARRY) << 7); if (value & 1) setcarry(); else clearcarry(); zerocalc(result); signcalc(result); putvalue(result); } static void rti() { status = pull8(); value = pull16(); pc = value; } static void rts() { value = pull16(); pc = value + 1; } static void sbc() { penaltyop = 1; value = getvalue() ^ 0x00FF; result = (uint16_t)a + value + (uint16_t)(status & FLAG_CARRY); carrycalc(result); zerocalc(result); overflowcalc(result, a, value); signcalc(result); #ifndef NES_CPU if (status & FLAG_DECIMAL) { clearcarry(); a -= 0x66; if ((a & 0x0F) > 0x09) { a += 0x06; } if ((a & 0xF0) > 0x90) { a += 0x60; setcarry(); } clockticks6502++; } #endif saveaccum(result); } static void sec() { setcarry(); } static void sed() { setdecimal(); } static void sei() { setinterrupt(); } static void sta() { putvalue(a); } static void stx() { putvalue(x); } static void sty() { putvalue(y); } static void tax() { x = a; zerocalc(x); signcalc(x); } static void tay() { y = a; zerocalc(y); signcalc(y); } static void tsx() { x = sp; zerocalc(x); signcalc(x); } static void txa() { a = x; zerocalc(a); signcalc(a); } static void txs() { sp = x; } static void tya() { a = y; zerocalc(a); signcalc(a); } //undocumented instructions #ifdef UNDOCUMENTED static void lax() { lda(); ldx(); } static void sax() { sta(); stx(); putvalue(a & x); if (penaltyop && penaltyaddr) clockticks6502--; } static void dcp() { dec(); cmp(); if (penaltyop && penaltyaddr) clockticks6502--; } static void isb() { inc(); sbc(); if (penaltyop && penaltyaddr) clockticks6502--; } static void slo() { asl(); ora(); if (penaltyop && penaltyaddr) clockticks6502--; } static void rla() { rol(); and(); if (penaltyop && penaltyaddr) clockticks6502--; } static void sre() { lsr(); eor(); if (penaltyop && penaltyaddr) clockticks6502--; } static void rra() { ror(); adc(); if (penaltyop && penaltyaddr) clockticks6502--; } #else #define lax nop #define sax nop #define dcp nop #define isb nop #define slo nop #define rla nop #define sre nop #define rra nop #endif static void (*addrtable[256])() = { /* | 0 | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | A | B | C | D | E | F | */ /* 0 */ imp, indx, imp, indx, zp, zp, zp, zp, imp, imm, acc, imm, abso, abso, abso, abso, /* 0 */ /* 1 */ rel, indy, imp, indy, zpx, zpx, zpx, zpx, imp, absy, imp, absy, absx, absx, absx, absx, /* 1 */ /* 2 */ abso, indx, imp, indx, zp, zp, zp, zp, imp, imm, acc, imm, abso, abso, abso, abso, /* 2 */ /* 3 */ rel, indy, imp, indy, zpx, zpx, zpx, zpx, imp, absy, imp, absy, absx, absx, absx, absx, /* 3 */ /* 4 */ imp, indx, imp, indx, zp, zp, zp, zp, imp, imm, acc, imm, abso, abso, abso, abso, /* 4 */ /* 5 */ rel, indy, imp, indy, zpx, zpx, zpx, zpx, imp, absy, imp, absy, absx, absx, absx, absx, /* 5 */ /* 6 */ imp, indx, imp, indx, zp, zp, zp, zp, imp, imm, acc, imm, ind, abso, abso, abso, /* 6 */ /* 7 */ rel, indy, imp, indy, zpx, zpx, zpx, zpx, imp, absy, imp, absy, absx, absx, absx, absx, /* 7 */ /* 8 */ imm, indx, imm, indx, zp, zp, zp, zp, imp, imm, imp, imm, abso, abso, abso, abso, /* 8 */ /* 9 */ rel, indy, imp, indy, zpx, zpx, zpy, zpy, imp, absy, imp, absy, absx, absx, absy, absy, /* 9 */ /* A */ imm, indx, imm, indx, zp, zp, zp, zp, imp, imm, imp, imm, abso, abso, abso, abso, /* A */ /* B */ rel, indy, imp, indy, zpx, zpx, zpy, zpy, imp, absy, imp, absy, absx, absx, absy, absy, /* B */ /* C */ imm, indx, imm, indx, zp, zp, zp, zp, imp, imm, imp, imm, abso, abso, abso, abso, /* C */ /* D */ rel, indy, imp, indy, zpx, zpx, zpx, zpx, imp, absy, imp, absy, absx, absx, absx, absx, /* D */ /* E */ imm, indx, imm, indx, zp, zp, zp, zp, imp, imm, imp, imm, abso, abso, abso, abso, /* E */ /* F */ rel, indy, imp, indy, zpx, zpx, zpx, zpx, imp, absy, imp, absy, absx, absx, absx, absx /* F */ }; static void (*optable[256])() = { /* | 0 | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | A | B | C | D | E | F | */ /* 0 */ brk, ora, nop, slo, nop, ora, asl, slo, php, ora, asl, nop, nop, ora, asl, slo, /* 0 */ /* 1 */ bpl, ora, nop, slo, nop, ora, asl, slo, clc, ora, nop, slo, nop, ora, asl, slo, /* 1 */ /* 2 */ jsr, and, nop, rla, bit, and, rol, rla, plp, and, rol, nop, bit, and, rol, rla, /* 2 */ /* 3 */ bmi, and, nop, rla, nop, and, rol, rla, sec, and, nop, rla, nop, and, rol, rla, /* 3 */ /* 4 */ rti, eor, nop, sre, nop, eor, lsr, sre, pha, eor, lsr, nop, jmp, eor, lsr, sre, /* 4 */ /* 5 */ bvc, eor, nop, sre, nop, eor, lsr, sre, cli, eor, nop, sre, nop, eor, lsr, sre, /* 5 */ /* 6 */ rts, adc, nop, rra, nop, adc, ror, rra, pla, adc, ror, nop, jmp, adc, ror, rra, /* 6 */ /* 7 */ bvs, adc, nop, rra, nop, adc, ror, rra, sei, adc, nop, rra, nop, adc, ror, rra, /* 7 */ /* 8 */ nop, sta, nop, sax, sty, sta, stx, sax, dey, nop, txa, nop, sty, sta, stx, sax, /* 8 */ /* 9 */ bcc, sta, nop, nop, sty, sta, stx, sax, tya, sta, txs, nop, nop, sta, nop, nop, /* 9 */ /* A */ ldy, lda, ldx, lax, ldy, lda, ldx, lax, tay, lda, tax, nop, ldy, lda, ldx, lax, /* A */ /* B */ bcs, lda, nop, lax, ldy, lda, ldx, lax, clv, lda, tsx, lax, ldy, lda, ldx, lax, /* B */ /* C */ cpy, cmp, nop, dcp, cpy, cmp, dec, dcp, iny, cmp, dex, nop, cpy, cmp, dec, dcp, /* C */ /* D */ bne, cmp, nop, dcp, nop, cmp, dec, dcp, cld, cmp, nop, dcp, nop, cmp, dec, dcp, /* D */ /* E */ cpx, sbc, nop, isb, cpx, sbc, inc, isb, inx, sbc, nop, sbc, cpx, sbc, inc, isb, /* E */ /* F */ beq, sbc, nop, isb, nop, sbc, inc, isb, sed, sbc, nop, isb, nop, sbc, inc, isb /* F */ }; static const uint32_t ticktable[256] = { /* | 0 | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | A | B | C | D | E | F | */ /* 0 */ 7, 6, 2, 8, 3, 3, 5, 5, 3, 2, 2, 2, 4, 4, 6, 6, /* 0 */ /* 1 */ 2, 5, 2, 8, 4, 4, 6, 6, 2, 4, 2, 7, 4, 4, 7, 7, /* 1 */ /* 2 */ 6, 6, 2, 8, 3, 3, 5, 5, 4, 2, 2, 2, 4, 4, 6, 6, /* 2 */ /* 3 */ 2, 5, 2, 8, 4, 4, 6, 6, 2, 4, 2, 7, 4, 4, 7, 7, /* 3 */ /* 4 */ 6, 6, 2, 8, 3, 3, 5, 5, 3, 2, 2, 2, 3, 4, 6, 6, /* 4 */ /* 5 */ 2, 5, 2, 8, 4, 4, 6, 6, 2, 4, 2, 7, 4, 4, 7, 7, /* 5 */ /* 6 */ 6, 6, 2, 8, 3, 3, 5, 5, 4, 2, 2, 2, 5, 4, 6, 6, /* 6 */ /* 7 */ 2, 5, 2, 8, 4, 4, 6, 6, 2, 4, 2, 7, 4, 4, 7, 7, /* 7 */ /* 8 */ 2, 6, 2, 6, 3, 3, 3, 3, 2, 2, 2, 2, 4, 4, 4, 4, /* 8 */ /* 9 */ 2, 6, 2, 6, 4, 4, 4, 4, 2, 5, 2, 5, 5, 5, 5, 5, /* 9 */ /* A */ 2, 6, 2, 6, 3, 3, 3, 3, 2, 2, 2, 2, 4, 4, 4, 4, /* A */ /* B */ 2, 5, 2, 5, 4, 4, 4, 4, 2, 4, 2, 4, 4, 4, 4, 4, /* B */ /* C */ 2, 6, 2, 8, 3, 3, 5, 5, 2, 2, 2, 2, 4, 4, 6, 6, /* C */ /* D */ 2, 5, 2, 8, 4, 4, 6, 6, 2, 4, 2, 7, 4, 4, 7, 7, /* D */ /* E */ 2, 6, 2, 8, 3, 3, 5, 5, 2, 2, 2, 2, 4, 4, 6, 6, /* E */ /* F */ 2, 5, 2, 8, 4, 4, 6, 6, 2, 4, 2, 7, 4, 4, 7, 7 /* F */ }; void nmi6502() { push16(pc); push8(status); status |= FLAG_INTERRUPT; pc = (uint16_t)read6502(0xFFFA) | ((uint16_t)read6502(0xFFFB) << 8); } void irq6502() { push16(pc); push8(status); status |= FLAG_INTERRUPT; pc = (uint16_t)read6502(0xFFFE) | ((uint16_t)read6502(0xFFFF) << 8); } uint8_t callexternal = 0; void (*loopexternal)(); void exec6502(uint32_t tickcount) { clockgoal6502 += tickcount; while (clockticks6502 < clockgoal6502) { opcode = read6502(pc++); penaltyop = 0; penaltyaddr = 0; (*addrtable[opcode])(); (*optable[opcode])(); clockticks6502 += ticktable[opcode]; if (penaltyop && penaltyaddr) clockticks6502++; instructions++; if (callexternal) (*loopexternal)(); } } void step6502() { opcode = read6502(pc++); penaltyop = 0; penaltyaddr = 0; (*addrtable[opcode])(); (*optable[opcode])(); clockticks6502 += ticktable[opcode]; if (penaltyop && penaltyaddr) clockticks6502++; clockgoal6502 = clockticks6502; instructions++; if (callexternal) (*loopexternal)(); } void hookexternal(void *funcptr) { if (funcptr != (void *)NULL) { loopexternal = funcptr; callexternal = 1; } else callexternal = 0; } 

Commentaires

• FYI, si vous utilisez la méthode markdown de marquage du code (indenté de quatre espaces), il sera dans une zone de défilement de la taille dun écran; et vous navez ‘ pas besoin de html-ize < balises entre crochets >. … Mais pour cette réponse, je pense en fait que ‘ est meilleur tel quel. En tant quimplémentation de référence, il met lespace nécessaire à une très bonne utilisation. … Si / quand plus de réponses arrivent, vous voudrez peut-être passer au retrait de 4 espaces pour quil ne domine pas ‘ la page. 0,02 $ … Jadore la question … +1 +1 +1! Je ‘ travaille sur le mien, ne ‘ t vous inquiétez pas! 🙂

Un émulateur MOS 6502 dans Haskell. Les fonctionnalités incluent:

• Implémentation précise du bit, y compris la gestion subtile du registre P et le retour à la page pendant lindexation et lindirection
• E / S mappées en mémoire, avec détection de boucle de rotation (donc le processeur hôte ne peg en attendant lentrée)
• détection darrêt (sauts / branchements sur soi)
• CPU implémentée en exactement 200 lignes & 6502 caractères de code
• Limplémentation du processeur est une pure monade détat

Ceci est une version quelque peu golfique dune implémentation complète (avec plus de fonctionnalités) que jai faite pour ce défi que jai » Je publierai plus tard. Malgré le golf, le code est toujours simple. La seule fonctionnalité manquante connue est le mode BCD (à venir …)

Exécute le code ehBASIC:

& ghc -O2 -o z6502min -Wall -fwarn-tabs -fno-warn-missing-signatures Z6502.hs [1 of 1] Compiling Main ( Z6502.hs, Z6502.o ) Z6502.hs:173:1: Warning: Defined but not used: `nmi" Z6502.hs:174:1: Warning: Defined but not used: `irq" Linking z6502min ... & ./z6502min ehbasic.bin 6502 EhBASIC [C]old/[W]arm ? Memory size ? 48383 Bytes free Enhanced BASIC 2.22 Ready PRINT "Hello World" Hello World Ready 10 FOR I = 1 TO 10 20 FOR J = 1 TO I 30 PRINT J; 40 NEXT J 50 PRINT 60 NEXT I RUN 1 1 2 1 2 3 1 2 3 4 1 2 3 4 5 1 2 3 4 5 6 1 2 3 4 5 6 7 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 9 1 2 3 4 5 6 7 8 9 10 Ready 

Et le code, à moins de 300 lignes au total:

-- Z6502: a 6502 emulator -- by Mark Lentczner module Main (main) where import Control.Applicative import Control.Monad import Control.Monad.State.Strict import Data.Bits import qualified Data.ByteString as B import Data.List import qualified Data.Vector as V import qualified Data.Vector.Unboxed as VU import Data.Word import System.Environment import System.IO {- === CPU: 200 lines, 6502 characters === -} type Addr = Word16 toAd = fromIntegral :: Int -> Addr addr :: Word8 -> Word8 -> Addr addr lo hi = fromIntegral hi `shiftL` 8 .|. fromIntegral lo lohi ad = (fromIntegral ad, fromIntegral $ ad `shiftR` 8) zeroPage v = addr v 0 index ad idx = ad + fromIntegral (idx :: Word8) relativeAddr ad off = index ad off - if off > 0x7f then 256 else 0 data Page = Missing | ROM !B.ByteString | RAM !(VU.Vector Word8) type Memory = V.Vector Page emptyMemory = V.replicate 256 Missing fetchByte ad mv = case mv V.! hi of ROM bs -> B.index bs lo RAM vs -> vs VU.! lo _ -> 0 where (hi,lo) = fromIntegral ad `divMod` 256 storeByte ad v mv = case mv V.! hi of RAM vs -> mv V.// [(hi, RAM $ vs VU.// [(lo, v)])] _ -> mv where (hi,lo) = fromIntegral ad `divMod` 256 data S = S { rA, rX, rY, rP, rS :: !Word8, rPC :: !Addr , mem :: !Memory, busR,busW :: Maybe Addr } powerOnState = S 0 0 0 0 0 0 emptyMemory Nothing Nothing [bitN, bitV, bitX, bitB, bitD, bitI, bitZ, bitC] = [7,6..0] toBit b t v = (if t then setBit else clearBit) v b toZ v = toBit bitZ (v == 0) toZN v = toBit bitZ (v == 0) . toBit bitN (testBit v 7) to67 v = toBit bitV (testBit v 6) . toBit bitN (testBit v 7) setZN v = modify $ \s -> s { rP = toZN v $ rP s } setAZN v = modify $ \s -> s { rA = v, rP=toZN v $ rP s } setXZN v = modify $ \s -> s { rX = v, rP=toZN v $ rP s } setYZN v = modify $ \s -> s { rY = v, rP=toZN v $ rP s } setZVNbit (a,v) = modify $ \s -> s { rP = toZ (a .&. v) $ to67 v $ rP s } setACZVN (c,v,a) = modify $ \s -> s { rA = a, rP = toBit bitC c $ toBit bitV v $ toZN a $ rP s } setCZN (c,v) = modify $ \s -> s { rP = toBit bitC c $ toZN v $ rP s } fetch a = state $ \s -> (fetchByte a $ mem s, s { busR = Just a }) fetchIndirectAddr a0 = do m <- gets mem let (lo,hi) = lohi a0 a1 = addr (lo+1) hi bLo = fetchByte a0 m bHi = fetchByte a1 m return $ addr bLo bHi store a v = modify $ \s -> s { mem = storeByte a v $ mem s, busW = Just a } clearBus = modify $ \s -> s { busR = Nothing, busW = Nothing } nextPC = state $ \s -> (rPC s, s { rPC = rPC s + 1 }) fetchPC = nextPC >>= \a -> gets mem >>= return . fetchByte a adjSP n m = state $ \s -> (addr (rS s + m) 1, s { rS = rS s + n }) push v = adjSP (-1) 0 >>= flip store v pull = adjSP 1 1 >>= fetch pushAddr a = let (lo, hi) = lohi a in push hi >> push lo pullAddr = addr <$> pull <*> pull pushP fromSW = gets rP >>= push . toBit bitX True . toBit bitB fromSW pullP = pull >>= \v -> modify $ \s -> s { rP = v .&. 0xCF } indexX a = gets rX >>= return . index a indexY a = gets rY >>= return . index a aImm=nextPC aZero=zeroPage<$>fetchPC aZeroX=zeroPage<$>((+)<$>fetchPC<*>gets rX) aZeroY=zeroPage<$>((+)<$>fetchPC<*>gets rY) aRel=flip relativeAddr<$>fetchPC<*>gets rPC aAbs=addr<$>fetchPC<*>fetchPC aAbsX=aAbs>>=indexX aAbsY=aAbs>>=indexY aInd=aAbs>>=fetchIndirectAddr aIndIdx=aZeroX>>=fetchIndirectAddr aIdxInd=aZero>>=fetchIndirectAddr>>=indexY decode = V.fromList $ concat $ transpose [[iBRK,iBPL,iJSR&aAbs,iBMI,iRTI,iBVC,iRTS,iBVS ,iErr,iBCC,iLDY&aImm,iBCS,iCPY&aImm,iBNE,iCPX&aImm,iBEQ] ,cAlu aIndIdx aIdxInd ,cErr//(10,iLDX&aImm) ,cErr ,[iErr,iErr,iBIT&aZero,iErr,iErr,iErr,iErr,iErr ,iSTY&aZero,iSTY&aZeroX,iLDY&aZero,iLDY&aZeroX,iCPY&aZero,iErr,iCPX&aZero,iErr] ,cAlu aZero aZeroX ,cBit aZero aZeroX//(9,iSTX&aZeroY)//(11,iLDX&aZeroY) ,cErr ,[iPHP,iCLC,iPLP,iSEC,iPHA,iCLI,iPLA,iSEI,iDEY,iTYA,iTAY,iCLV,iINY,iCLD,iINX,iSED] ,cAlu aImm aAbsY//(8,iErr) ,[iASLa,iErr,iROLa,iErr,iLSRa,iErr,iRORa,iErr ,iTXA,iTXS,iTAX,iTSX,iDEX,iErr,iNOP,iErr ] ,cErr ,[iErr,iErr,iBIT&aAbs,iErr,iJMP&aAbs,iErr,iJMP&aInd,iErr ,iSTY&aAbs,iErr,iLDY&aAbs,iLDY&aAbsX,iCPY&aAbs,iErr,iCPX&aAbs,iErr] ,cAlu aAbs aAbsX ,cBit aAbs aAbsX//(9,iErr)//(11,iLDX&aAbsY) ,cErr ] cAlt is e o = is >>= (\i->[i&e,i&o]) cAlu = cAlt [iORA,iAND,iEOR,iADC,iSTA,iLDA,iCMP,iSBC] cBit = cAlt [iASL,iROL,iLSR,iROR,iSTX,iLDX,iDEC,iINC] cErr = replicate 16 iErr is//(n,j) = let (f,_:h) = splitAt n is in f++j:h i&a=a>>=i loadIns l a = fetch a >>= l storeIns f a = f >>= store a aluIns set op ad = do v <- fetch ad a <- gets rA set $ op a v modIns op a = fetch a >>= op >>= store a modAccIns op = gets rA >>= op >>= \v -> modify $ \s -> s { rA = v } stIns b op = modify $ \s -> s { rP = op (rP s) b } jump a = modify $ \s -> s { rPC = a } brIns b t = do a <- aRel p <- gets rP when (testBit p b == t) $ jump a adcOp a b cIn = (cOut, v, s) where h = b + (if cIn then 1 else 0) s = a + h cOut = h < b || s < a v = testBit (a `xor` s .&. b `xor` s) 7 sbcOp a b cIn = adcOp a (complement b) cIn carryOp f = gets rP >>= setACZVN . f . flip testBit bitC cmpOp a b = (a >= b, a - b) shiftOp shifter isRot inBit outBit v = do s <- get let newC = testBit v outBit bitIn = toBit inBit $ isRot && testBit (rP s) bitC v" = bitIn $ shifter v 1 put s { rP = toBit bitC newC $ toZN v" $ rP s } return v" vector a = fetchIndirectAddr a >>= jump interrupt isBrk pcOffset a = do gets rPC >>= pushAddr . flip index pcOffset pushP isBrk iSEI vector a reset = vector $ toAd 0xFFFC nmi = interrupt False 0 $ toAd 0xFFFA irq = interrupt False 0 $ toAd 0xFFFE [iORA,iAND,iEOR]=aluIns setAZN<$>[(.|.),(.&.),xor] [iADC,iSBC]=aluIns carryOp<$>[adcOp,sbcOp] iSTA=storeIns$gets rA iLDA=loadIns setAZN iCMP=aluIns setCZN cmpOp [iSTX,iSTY]=storeIns.gets<$>[rX,rY] [iLDX,iLDY]=loadIns<$>[setXZN,setYZN] [iCPX,iCPY]=(\r a->gets r>>= \v->fetch a>>=setCZN.cmpOp v)<$>[rX,rY] [iDEC,iINC]=modIns.(\i v->setZN(v+i)>>return(v+i))<$>[-1,1] [iDEX,iINX]=(gets rX>>=).(setXZN.).(+)<$>[-1,1] [iDEY,iINY]=(gets rY>>=).(setYZN.).(+)<$>[-1,1] shOps=[shiftOp d r b(7-b)|(d,b)<-[(shiftL,0),(shiftR,7)],r<-[False,True]] [iASL,iROL,iLSR,iROR]=modIns<$>shOps [iASLa,iROLa,iLSRa,iRORa]=modAccIns<$>shOps iBIT=aluIns setZVNbit(,) iJMP=jump [iBPL,iBMI,iBVC,iBVS,iBCC,iBCS,iBNE,iBEQ]=brIns<$>[bitN,bitV,bitC,bitZ]<*>[False,True] [iCLC,iSEC,iCLI,iSEI,iCLV,_,iCLD,iSED]=stIns<$>[bitC,bitI,bitV,bitD]<*>[clearBit,setBit] iBRK=interrupt True 1 $ toAd 0xFFFE iJSR a=gets rPC>>=pushAddr.(-1+)>>jump a iRTI=iPLP>>pullAddr>>=jump iRTS=pullAddr>>=jump.(1+) iPHP=pushP True iPLP=pullP iPHA=gets rA>>=push iPLA=pull>>=setAZN iNOP=return () [iTAX,iTAY]=(gets rA>>=)<$>[setXZN,setYZN] [iTXA,iTYA]=(>>=setAZN).gets<$>[rX,rY] iTXS=modify $ \s -> s { rS=rX s } iTSX=gets rS>>=setXZN iErr=gets rPC>>=jump.(-1+) executeOne = clearBus >> fetchPC >>= (decode V.!) . fromIntegral {- === END OF CPU === -} {- === MOTHERBOARD === -} buildMemory rom = loadRAM 0xF0 1 $ loadRAM 0x00 ramSize $ loadROM romStart rom $ emptyMemory where ramSize = 256 - (B.length rom `div` 256) romStart = fromIntegral ramSize loadRAM p0 n = (V.// zip [p0..] (map RAM $ replicate n ramPage)) ramPage = VU.replicate 256 0 loadROM p0 bs = (V.// zip [p0..] (map ROM $ romPages bs)) romPages b = case B.length b of l | l == 0 -> [] | l < 256 -> [b `B.append` B.replicate (256 - l) 0] | l == 256 -> [b] | otherwise -> let (b0,bn) = B.splitAt 256 b in b0 : romPages bn main = getArgs >>= go where go [romFile] = B.readFile romFile >>= exec . buildState . buildMemory go _ = putStrLn "agument should be a single ROM file" buildState m = execState reset (powerOnState { mem = m }) exec s0 = do stopIO <- startIO loop (0 :: Int) s0 stopIO loop n s = do let pcsp = (rPC s, rS s) (n",s") <- processIO n (execState executeOne s) let pcsp" = (rPC s", rS s") if pcsp /= pcsp" then (loop $! n") $! s" else do putStrLn $ "Execution snagged at " ++ show (fst pcsp") startIO = do ibuf <- hGetBuffering stdin obuf <- hGetBuffering stdout iecho <- hGetEcho stdin hSetBuffering stdin NoBuffering hSetBuffering stdout NoBuffering hSetEcho stdin False return $ do hSetEcho stdin iecho hSetBuffering stdin ibuf hSetBuffering stdout obuf putStr "\n\n" processIO n s = do when (busW s == Just outPortAddr) $ do let c = fetchByte outPortAddr $ mem s when (c /= 0) $ hPutChar stdout $ toEnum $ fromIntegral c if (busR s == Just inPortAddr) then do r <- if n < 16 then hWaitForInput stdin 50 else hReady stdin c <- if r then (fromIntegral . fromEnum) <$> hGetChar stdin else return 0 let c" = if c == 0xA then 0xD else c let s" = s { mem = storeByte inPortAddr c" $ mem s } return (0,s") else return (n+1,s) inPortAddr = toAd 0xF004 outPortAddr = toAd 0xF001 

Commentaires

• Bon travail! Très petit. Je ne ‘ ne connais pas Haskell, je devrais peut-être apprendre. Jadore le fait que s 6502 caractères. 🙂

Pour toute personne intéressée, jai pensé que je partagerais mon implémentation du 6502 en C #. Comme avec dautres articles ici, il est complètement non golfé mais est une implémentation complète de fonctionnalité.

• Prend en charge NMOS et CMOS
• Inclut plusieurs programmes de test, y compris le test AllSuite ci-dessus en tant que tests unitaires.
• Supporte BCD

Jai commencé ce projet en créant une feuille de calcul dinstructions lorsque jai découvert le CPU pour la première fois. Jai réalisé que je pouvais utiliser cette feuille de calcul pour me sauver un peu de frappe. Jai transformé cela en un tableau de fichiers texte que lémulateur charge pour aider à compter les cycles et pour faciliter la sortie du démontage.

Lensemble du projet est disponible sur Github https://github.com/amensch/e6502

/* * e6502: A complete 6502 CPU emulator. * Copyright 2016 Adam Mensch */ using System; namespace e6502CPU { public enum e6502Type { CMOS, NMOS }; public class e6502 { // Main Register public byte A; // Index Registers public byte X; public byte Y; // Program Counter public ushort PC; // Stack Pointer // Memory location is hard coded to 0x01xx // Stack is descending (decrement on push, increment on pop) // 6502 is an empty stack so SP points to where next value is stored public byte SP; // Status Registers (in order bit 7 to 0) public bool NF; // negative flag (N) public bool VF; // overflow flag (V) // bit 5 is unused // bit 4 is the break flag however it is not a physical flag in the CPU public bool DF; // binary coded decimal flag (D) public bool IF; // interrupt flag (I) public bool ZF; // zero flag (Z) public bool CF; // carry flag (C) // RAM - 16 bit address bus means 64KB of addressable memory public byte[] memory; // List of op codes and their attributes private OpCodeTable _opCodeTable; // The current opcode private OpCodeRecord _currentOP; // Clock cycles to adjust due to page boundaries being crossed, branches taken, or NMOS/CMOS differences private int _extraCycles; // Flag for hardware interrupt (IRQ) public bool IRQWaiting { get; set; } // Flag for non maskable interrupt (NMI) public bool NMIWaiting { get; set; } public e6502Type _cpuType { get; set; } public e6502(e6502Type type) { memory = new byte[0x10000]; _opCodeTable = new OpCodeTable(); // Set these on instantiation so they are known values when using this object in testing. // Real programs should explicitly load these values before using them. A = 0; X = 0; Y = 0; SP = 0; PC = 0; NF = false; VF = false; DF = false; IF = true; ZF = false; CF = false; NMIWaiting = false; IRQWaiting = false; _cpuType = type; } public void Boot() { // On reset the addresses 0xfffc and 0xfffd are read and PC is loaded with this value. // It is expected that the initial program loaded will have these values set to something. // Most 6502 systems contain ROM in the upper region (around 0xe000-0xffff) PC = GetWordFromMemory(0xfffc); // interrupt disabled is set on powerup IF = true; NMIWaiting = false; IRQWaiting = false; } public void LoadProgram(ushort startingAddress, byte[] program) { program.CopyTo(memory, startingAddress); PC = startingAddress; } public string DasmNextInstruction() { OpCodeRecord oprec = _opCodeTable.OpCodes[ memory[PC] ]; if (oprec.Bytes == 3) return oprec.Dasm( GetImmWord() ); else return oprec.Dasm( GetImmByte() ); } // returns # of clock cycles needed to execute the instruction public int ExecuteNext() { _extraCycles = 0; // Check for non maskable interrupt (has higher priority over IRQ) if (NMIWaiting) { DoIRQ(0xfffa); NMIWaiting = false; _extraCycles += 6; } // Check for hardware interrupt, if enabled else if (!IF) { if(IRQWaiting) { DoIRQ(0xfffe); IRQWaiting = false; _extraCycles += 6; } } _currentOP = _opCodeTable.OpCodes[memory[PC]]; ExecuteInstruction(); return _currentOP.Cycles + _extraCycles; } private void ExecuteInstruction() { int result; int oper = GetOperand(_currentOP.AddressMode); switch (_currentOP.OpCode) { // ADC - add memory to accumulator with carry // A+M+C -> A,C (NZCV) case 0x61: case 0x65: case 0x69: case 0x6d: case 0x71: case 0x72: case 0x75: case 0x79: case 0x7d: if (DF) { result = HexToBCD(A) + HexToBCD((byte)oper); if (CF) result++; CF = (result > 99); if (result > 99 ) { result -= 100; } ZF = (result == 0); // convert decimal result to hex BCD result A = BCDToHex(result); // Unlike ZF and CF, the NF flag represents the MSB after conversion // to BCD. NF = (A > 0x7f); // extra clock cycle on CMOS in decimal mode if (_cpuType == e6502Type.CMOS) _extraCycles++; } else { ADC((byte)oper); } PC += _currentOP.Bytes; break; // AND - and memory with accumulator // A AND M -> A (NZ) case 0x21: case 0x25: case 0x29: case 0x2d: case 0x31: case 0x32: case 0x35: case 0x39: case 0x3d: result = A & oper; NF = ((result & 0x80) == 0x80); ZF = ((result & 0xff) == 0x00); A = (byte)result; PC += _currentOP.Bytes; break; // ASL - shift left one bit (NZC) // C <- (76543210) <- 0 case 0x06: case 0x16: case 0x0a: case 0x0e: case 0x1e: // On 65C02 (abs,X) takes one less clock cycle (but still add back 1 if page boundary crossed) if (_currentOP.OpCode == 0x1e && _cpuType == e6502Type.CMOS) _extraCycles--; // shift bit 7 into carry CF = (oper >= 0x80); // shift operand result = oper << 1; NF = ((result & 0x80) == 0x80); ZF = ((result & 0xff) == 0x00); SaveOperand(_currentOP.AddressMode, result); PC += _currentOP.Bytes; break; // BBRx - test bit in memory (no flags) // Test the zero page location and branch of the specified bit is clear // These instructions are only available on Rockwell and WDC 65C02 chips. // Number of clock cycles is the same regardless if the branch is taken. case 0x0f: case 0x1f: case 0x2f: case 0x3f: case 0x4f: case 0x5f: case 0x6f: case 0x7f: // upper nibble specifies the bit to check byte check_bit = (byte)(_currentOP.OpCode >> 4); byte check_value = 0x01; for( int ii=0; ii < check_bit; ii++) { check_value = (byte)(check_value << 1); } // if the specified bit is 0 then branch byte offset = memory[PC + 2]; PC += _currentOP.Bytes; if ((oper & check_value) == 0x00) PC += offset; break; // BBSx - test bit in memory (no flags) // Test the zero page location and branch of the specified bit is set // These instructions are only available on Rockwell and WDC 65C02 chips. // Number of clock cycles is the same regardless if the branch is taken. case 0x8f: case 0x9f: case 0xaf: case 0xbf: case 0xcf: case 0xdf: case 0xef: case 0xff: // upper nibble specifies the bit to check (but ignore bit 7) check_bit = (byte)((_currentOP.OpCode & 0x70) >> 4); check_value = 0x01; for (int ii = 0; ii < check_bit; ii++) { check_value = (byte)(check_value << 1); } // if the specified bit is 1 then branch offset = memory[PC + 2]; PC += _currentOP.Bytes; if ((oper & check_value) == check_value) PC += offset; break; // BCC - branch on carry clear case 0x90: PC += _currentOP.Bytes; CheckBranch(!CF, oper); break; // BCS - branch on carry set case 0xb0: PC += _currentOP.Bytes; CheckBranch(CF, oper); break; // BEQ - branch on zero case 0xf0: PC += _currentOP.Bytes; CheckBranch(ZF, oper); break; // BIT - test bits in memory with accumulator (NZV) // bits 7 and 6 of oper are transferred to bits 7 and 6 of conditional register (N and V) // the zero flag is set to the result of oper AND accumulator case 0x24: case 0x2c: // added by 65C02 case 0x34: case 0x3c: case 0x89: result = A & oper; // The WDC programming manual for 65C02 indicates NV are unaffected in immediate mode. // The extended op code test program reflects this. if (_currentOP.AddressMode != AddressModes.Immediate) { NF = ((oper & 0x80) == 0x80); VF = ((oper & 0x40) == 0x40); } ZF = ((result & 0xff) == 0x00); PC += _currentOP.Bytes; break; // BMI - branch on negative case 0x30: PC += _currentOP.Bytes; CheckBranch(NF, oper); break; // BNE - branch on non zero case 0xd0: PC += _currentOP.Bytes; CheckBranch(!ZF, oper); break; // BPL - branch on non negative case 0x10: PC += _currentOP.Bytes; CheckBranch(!NF, oper); break; // BRA - unconditional branch to immediate address // NOTE: In OpcodeList.txt the number of clock cycles is one less than the documentation. // This is because CheckBranch() adds one when a branch is taken, which in this case is always. case 0x80: PC += _currentOP.Bytes; CheckBranch(true, oper); break; // BRK - force break (I) case 0x00: // This is a software interrupt (IRQ). These events happen in a specific order. // Processor adds two to the current PC PC += 2; // Call IRQ routine DoIRQ(0xfffe, true); // Whether or not the decimal flag is cleared depends on the type of 6502 CPU. // The CMOS 65C02 clears this flag but the NMOS 6502 does not. if( _cpuType == e6502Type.CMOS ) DF = false; break; // BVC - branch on overflow clear case 0x50: PC += _currentOP.Bytes; CheckBranch(!VF, oper); break; // BVS - branch on overflow set case 0x70: PC += _currentOP.Bytes; CheckBranch(VF, oper); break; // CLC - clear carry flag case 0x18: CF = false; PC += _currentOP.Bytes; break; // CLD - clear decimal mode case 0xd8: DF = false; PC += _currentOP.Bytes; break; // CLI - clear interrupt disable bit case 0x58: IF = false; PC += _currentOP.Bytes; break; // CLV - clear overflow flag case 0xb8: VF = false; PC += _currentOP.Bytes; break; // CMP - compare memory with accumulator (NZC) // CMP, CPX and CPY are unsigned comparisions case 0xc5: case 0xc9: case 0xc1: case 0xcd: case 0xd1: case 0xd2: case 0xd5: case 0xd9: case 0xdd: byte temp = (byte)(A - oper); CF = A >= (byte)oper; ZF = A == (byte)oper; NF = ((temp & 0x80) == 0x80); PC += _currentOP.Bytes; break; // CPX - compare memory and X (NZC) case 0xe0: case 0xe4: case 0xec: temp = (byte)(X - oper); CF = X >= (byte)oper; ZF = X == (byte)oper; NF = ((temp & 0x80) == 0x80); PC += _currentOP.Bytes; break; // CPY - compare memory and Y (NZC) case 0xc0: case 0xc4: case 0xcc: temp = (byte)(Y - oper); CF = Y >= (byte)oper; ZF = Y == (byte)oper; NF = ((temp & 0x80) == 0x80); PC += _currentOP.Bytes; break; // DEC - decrement memory by 1 (NZ) case 0xc6: case 0xce: case 0xd6: case 0xde: // added by 65C02 case 0x3a: result = oper - 1; ZF = ((result & 0xff) == 0x00); NF = ((result & 0x80) == 0x80); SaveOperand(_currentOP.AddressMode, result); PC += _currentOP.Bytes; break; // DEX - decrement X by one (NZ) case 0xca: result = X - 1; ZF = ((result & 0xff) == 0x00); NF = ((result & 0x80) == 0x80); X = (byte)result; PC += _currentOP.Bytes; break; // DEY - decrement Y by one (NZ) case 0x88: result = Y - 1; ZF = ((result & 0xff) == 0x00); NF = ((result & 0x80) == 0x80); Y = (byte)result; PC += _currentOP.Bytes; break; // EOR - XOR memory with accumulator (NZ) case 0x41: case 0x45: case 0x49: case 0x4d: case 0x51: case 0x52: case 0x55: case 0x59: case 0x5d: result = A ^ (byte)oper; ZF = ((result & 0xff) == 0x00); NF = ((result & 0x80) == 0x80); A = (byte)result; PC += _currentOP.Bytes; break; // INC - increment memory by 1 (NZ) case 0xe6: case 0xee: case 0xf6: case 0xfe: // added by 65C02 case 0x1a: result = oper + 1; ZF = ((result & 0xff) == 0x00); NF = ((result & 0x80) == 0x80); SaveOperand(_currentOP.AddressMode, result); PC += _currentOP.Bytes; break; // INX - increment X by one (NZ) case 0xe8: result = X + 1; ZF = ((result & 0xff) == 0x00); NF = ((result & 0x80) == 0x80); X = (byte)result; PC += _currentOP.Bytes; break; // INY - increment Y by one (NZ) case 0xc8: result = Y + 1; ZF = ((result & 0xff) == 0x00); NF = ((result & 0x80) == 0x80); Y = (byte)result; PC += _currentOP.Bytes; break; // JMP - jump to new location (two byte immediate) case 0x4c: case 0x6c: // added for 65C02 case 0x7c: if (_currentOP.AddressMode == AddressModes.Absolute) { PC = GetImmWord(); } else if (_currentOP.AddressMode == AddressModes.Indirect) { PC = (ushort)(GetWordFromMemory(GetImmWord())); } else if( _currentOP.AddressMode == AddressModes.AbsoluteX) { PC = GetWordFromMemory((GetImmWord() + X)); } else { throw new InvalidOperationException("This address mode is invalid with the JMP instruction"); } // CMOS fixes a bug in this op code which results in an extra clock cycle if (_currentOP.OpCode == 0x6c && _cpuType == e6502Type.CMOS) _extraCycles++; break; // JSR - jump to new location and save return address case 0x20: // documentation says push PC+2 even though this is a 3 byte instruction // When pulled via RTS 1 is added to the result Push((ushort)(PC+2)); PC = GetImmWord(); break; // LDA - load accumulator with memory (NZ) case 0xa1: case 0xa5: case 0xa9: case 0xad: case 0xb1: case 0xb2: case 0xb5: case 0xb9: case 0xbd: A = (byte)oper; ZF = ((A & 0xff) == 0x00); NF = ((A & 0x80) == 0x80); PC += _currentOP.Bytes; break; // LDX - load index X with memory (NZ) case 0xa2: case 0xa6: case 0xae: case 0xb6: case 0xbe: X = (byte)oper; ZF = ((X & 0xff) == 0x00); NF = ((X & 0x80) == 0x80); PC += _currentOP.Bytes; break; // LDY - load index Y with memory (NZ) case 0xa0: case 0xa4: case 0xac: case 0xb4: case 0xbc: Y = (byte)oper; ZF = ((Y & 0xff) == 0x00); NF = ((Y & 0x80) == 0x80); PC += _currentOP.Bytes; break; // LSR - shift right one bit (NZC) // 0 -> (76543210) -> C case 0x46: case 0x4a: case 0x4e: case 0x56: case 0x5e: // On 65C02 (abs,X) takes one less clock cycle (but still add back 1 if page boundary crossed) if (_currentOP.OpCode == 0x5e && _cpuType == e6502Type.CMOS) _extraCycles--; // shift bit 0 into carry CF = ((oper & 0x01) == 0x01); // shift operand result = oper >> 1; ZF = ((result & 0xff) == 0x00); NF = ((result & 0x80) == 0x80); SaveOperand(_currentOP.AddressMode, result); PC += _currentOP.Bytes; break; // NOP - no operation case 0xea: PC += _currentOP.Bytes; break; // ORA - OR memory with accumulator (NZ) case 0x01: case 0x05: case 0x09: case 0x0d: case 0x11: case 0x12: case 0x15: case 0x19: case 0x1d: result = A | (byte)oper; ZF = ((result & 0xff) == 0x00); NF = ((result & 0x80) == 0x80); A = (byte)result; PC += _currentOP.Bytes; break; // PHA - push accumulator on stack case 0x48: Push(A); PC += _currentOP.Bytes; break; // PHP - push processor status on stack case 0x08: int sr = 0x00; if (NF) sr = sr | 0x80; if (VF) sr = sr | 0x40; sr = sr | 0x20; // bit 5 is always 1 sr = sr | 0x10; // bit 4 is always 1 for PHP if (DF) sr = sr | 0x08; if (IF) sr = sr | 0x04; if (ZF) sr = sr | 0x02; if (CF) sr = sr | 0x01; Push((byte)sr); PC += _currentOP.Bytes; break; // PHX - push X on stack case 0xda: Push(X); PC += _currentOP.Bytes; break; // PHY - push Y on stack case 0x5a: Push(Y); PC += _currentOP.Bytes; break; // PLA - pull accumulator from stack (NZ) case 0x68: A = PopByte(); NF = (A & 0x80) == 0x80; ZF = (A & 0xff) == 0x00; PC += _currentOP.Bytes; break; // PLP - pull status from stack case 0x28: sr = PopByte(); NF = (sr & 0x80) == 0x80; VF = (sr & 0x40) == 0x40; DF = (sr & 0x08) == 0x08; IF = (sr & 0x04) == 0x04; ZF = (sr & 0x02) == 0x02; CF = (sr & 0x01) == 0x01; PC += _currentOP.Bytes; break; // PLX - pull X from stack (NZ) case 0xfa: X = PopByte(); NF = (X & 0x80) == 0x80; ZF = (X & 0xff) == 0x00; PC += _currentOP.Bytes; break; // PLY - pull Y from stack (NZ) case 0x7a: Y = PopByte(); NF = (Y & 0x80) == 0x80; ZF = (Y & 0xff) == 0x00; PC += _currentOP.Bytes; break; // RMBx - clear bit in memory (no flags) // Clear the zero page location of the specified bit // These instructions are only available on Rockwell and WDC 65C02 chips. case 0x07: case 0x17: case 0x27: case 0x37: case 0x47: case 0x57: case 0x67: case 0x77: // upper nibble specifies the bit to check check_bit = (byte)(_currentOP.OpCode >> 4); check_value = 0x01; for (int ii = 0; ii < check_bit; ii++) { check_value = (byte)(check_value << 1); } check_value = (byte)~check_value; SaveOperand(_currentOP.AddressMode, oper & check_value); PC += _currentOP.Bytes; break; // SMBx - set bit in memory (no flags) // Set the zero page location of the specified bit // These instructions are only available on Rockwell and WDC 65C02 chips. case 0x87: case 0x97: case 0xa7: case 0xb7: case 0xc7: case 0xd7: case 0xe7: case 0xf7: // upper nibble specifies the bit to check (but ignore bit 7) check_bit = (byte)((_currentOP.OpCode & 0x70) >> 4); check_value = 0x01; for (int ii = 0; ii < check_bit; ii++) { check_value = (byte)(check_value << 1); } SaveOperand(_currentOP.AddressMode, oper | check_value); PC += _currentOP.Bytes; break; // ROL - rotate left one bit (NZC) // C <- 76543210 <- C case 0x26: case 0x2a: case 0x2e: case 0x36: case 0x3e: // On 65C02 (abs,X) takes one less clock cycle (but still add back 1 if page boundary crossed) if (_currentOP.OpCode == 0x3e && _cpuType == e6502Type.CMOS) _extraCycles--; // perserve existing cf value bool old_cf = CF; // shift bit 7 into carry flag CF = (oper >= 0x80); // shift operand result = oper << 1; // old carry flag goes to bit zero if (old_cf) result = result | 0x01; ZF = ((result & 0xff) == 0x00); NF = ((result & 0x80) == 0x80); SaveOperand(_currentOP.AddressMode, result); PC += _currentOP.Bytes; break; // ROR - rotate right one bit (NZC) // C -> 76543210 -> C case 0x66: case 0x6a: case 0x6e: case 0x76: case 0x7e: // On 65C02 (abs,X) takes one less clock cycle (but still add back 1 if page boundary crossed) if (_currentOP.OpCode == 0x7e && _cpuType == e6502Type.CMOS) _extraCycles--; // perserve existing cf value old_cf = CF; // shift bit 0 into carry flag CF = (oper & 0x01) == 0x01; // shift operand result = oper >> 1; // old carry flag goes to bit 7 if (old_cf) result = result | 0x80; ZF = ((result & 0xff) == 0x00); NF = ((result & 0x80) == 0x80); SaveOperand(_currentOP.AddressMode, result); PC += _currentOP.Bytes; break; // RTI - return from interrupt case 0x40: // pull SR sr = PopByte(); NF = (sr & 0x80) == 0x80; VF = (sr & 0x40) == 0x40; DF = (sr & 0x08) == 0x08; IF = (sr & 0x04) == 0x04; ZF = (sr & 0x02) == 0x02; CF = (sr & 0x01) == 0x01; // pull PC PC = PopWord(); break; // RTS - return from subroutine case 0x60: PC = (ushort)(PopWord() + 1); break; // SBC - subtract memory from accumulator with borrow (NZCV) // A-M-C -> A (NZCV) case 0xe1: case 0xe5: case 0xe9: case 0xed: case 0xf1: case 0xf2: case 0xf5: case 0xf9: case 0xfd: if (DF) { result = HexToBCD(A) - HexToBCD((byte)oper); if (!CF) result--; CF = (result >= 0); // BCD numbers wrap around when subtraction is negative if (result < 0) result += 100; ZF = (result == 0); A = BCDToHex(result); // Unlike ZF and CF, the NF flag represents the MSB after conversion // to BCD. NF = (A > 0x7f); // extra clock cycle on CMOS in decimal mode if (_cpuType == e6502Type.CMOS) _extraCycles++; } else { ADC((byte)~oper); } PC += _currentOP.Bytes; break; // SEC - set carry flag case 0x38: CF = true; PC += _currentOP.Bytes; break; // SED - set decimal mode case 0xf8: DF = true; PC += _currentOP.Bytes; break; // SEI - set interrupt disable bit case 0x78: IF = true; PC += _currentOP.Bytes; break; // STA - store accumulator in memory case 0x81: case 0x85: case 0x8d: case 0x91: case 0x92: case 0x95: case 0x99: case 0x9d: SaveOperand(_currentOP.AddressMode, A); PC += _currentOP.Bytes; break; // STX - store X in memory case 0x86: case 0x8e: case 0x96: SaveOperand(_currentOP.AddressMode, X); PC += _currentOP.Bytes; break; // STY - store Y in memory case 0x84: case 0x8c: case 0x94: SaveOperand(_currentOP.AddressMode, Y); PC += _currentOP.Bytes; break; // STZ - Store zero case 0x64: case 0x74: case 0x9c: case 0x9e: SaveOperand(_currentOP.AddressMode, 0); PC += _currentOP.Bytes; break; // TAX - transfer accumulator to X (NZ) case 0xaa: X = A; ZF = ((X & 0xff) == 0x00); NF = ((X & 0x80) == 0x80); PC += _currentOP.Bytes; break; // TAY - transfer accumulator to Y (NZ) case 0xa8: Y = A; ZF = ((Y & 0xff) == 0x00); NF = ((Y & 0x80) == 0x80); PC += _currentOP.Bytes; break; // TRB - test and reset bits (Z) // Perform bitwise AND between accumulator and contents of memory case 0x14: case 0x1c: SaveOperand(_currentOP.AddressMode, ~A & oper); ZF = (A & oper) == 0x00; PC += _currentOP.Bytes; break; // TSB - test and set bits (Z) // Perform bitwise AND between accumulator and contents of memory case 0x04: case 0x0c: SaveOperand(_currentOP.AddressMode, A | oper); ZF = (A & oper) == 0x00; PC += _currentOP.Bytes; break; // TSX - transfer SP to X (NZ) case 0xba: X = SP; ZF = ((X & 0xff) == 0x00); NF = ((X & 0x80) == 0x80); PC += _currentOP.Bytes; break; // TXA - transfer X to A (NZ) case 0x8a: A = X; ZF = ((A & 0xff) == 0x00); NF = ((A & 0x80) == 0x80); PC += _currentOP.Bytes; break; // TXS - transfer X to SP (no flags -- some online docs are incorrect) case 0x9a: SP = X; PC += _currentOP.Bytes; break; // TYA - transfer Y to A (NZ) case 0x98: A = Y; ZF = ((A & 0xff) == 0x00); NF = ((A & 0x80) == 0x80); PC += _currentOP.Bytes; break; // The original 6502 has undocumented and erratic behavior if // undocumented op codes are invoked. The 65C02 on the other hand // are guaranteed to be NOPs although they vary in number of bytes // and cycle counts. These NOPs are listed in the OpcodeList.txt file // so the proper number of clock cycles are used. // // Instructions STP (0xdb) and WAI (0xcb) will reach this case. // For now these are treated as a NOP. default: PC += _currentOP.Bytes; break; } } private int GetOperand(AddressModes mode) { int oper = 0; switch (mode) { // Accumulator mode uses the value in the accumulator case AddressModes.Accumulator: oper = A; break; // Retrieves the byte at the specified memory location case AddressModes.Absolute: oper = memory[ GetImmWord() ]; break; // Indexed absolute retrieves the byte at the specified memory location case AddressModes.AbsoluteX: ushort imm = GetImmWord(); ushort result = (ushort)(imm + X); if (_currentOP.CheckPageBoundary) { if ((imm & 0xff00) != (result & 0xff00)) _extraCycles += 1; } oper = memory[ result ]; break; case AddressModes.AbsoluteY: imm = GetImmWord(); result = (ushort)(imm + Y); if (_currentOP.CheckPageBoundary) { if ((imm & 0xff00) != (result & 0xff00)) _extraCycles += 1; } oper = memory[result]; break; // Immediate mode uses the next byte in the instruction directly. case AddressModes.Immediate: oper = GetImmByte(); break; // Implied or Implicit are single byte instructions that do not use // the next bytes for the operand. case AddressModes.Implied: break; // Indirect mode uses the absolute address to get another address. // The immediate word is a memory location from which to retrieve // the 16 bit operand. case AddressModes.Indirect: oper = GetWordFromMemory(GetImmWord()); break; // The indexed indirect modes uses the immediate byte rather than the // immediate word to get the memory location from which to retrieve // the 16 bit operand. This is a combination of ZeroPage indexed and Indirect. case AddressModes.XIndirect: /* * 1) fetch immediate byte * 2) add X to the byte * 3) obtain word from this zero page address * 4) return the byte located at the address specified by the word */ oper = memory[GetWordFromMemory( (byte)(GetImmByte() + X))]; break; // The Indirect Indexed works a bit differently than above. // The Y register is added *after* the deferencing instead of before. case AddressModes.IndirectY: /* 1) Fetch the address (word) at the immediate zero page location 2) Add Y to obtain the final target address 3)Load the byte at this address */ ushort addr = GetWordFromMemory(GetImmByte()); oper = memory[addr + Y]; if (_currentOP.CheckPageBoundary) { if ((oper & 0xff00) != (addr & 0xff00)) _extraCycles++; } break; // Relative is used for branching, the immediate value is a // signed 8 bit value and used to offset the current PC. case AddressModes.Relative: oper = SignExtend(GetImmByte()); break; // Zero Page mode is a fast way of accessing the first 256 bytes of memory. // Best programming practice is to place your variables in 0x00-0xff. // Retrieve the byte at the indicated memory location. case AddressModes.ZeroPage: oper = memory[GetImmByte()]; break; case AddressModes.ZeroPageX: oper = memory[(GetImmByte() + X) & 0xff]; break; case AddressModes.ZeroPageY: oper = memory[(GetImmByte() + Y) & 0xff]; break; // this mode is from the 65C02 extended set // works like ZeroPageY when Y=0 case AddressModes.ZeroPage0: oper = memory[GetWordFromMemory((GetImmByte()) & 0xff)]; break; // for this mode do the same thing as ZeroPage case AddressModes.BranchExt: oper = memory[GetImmByte()]; break; default: break; } return oper; } private void SaveOperand(AddressModes mode, int data) { switch (mode) { // Accumulator mode uses the value in the accumulator case AddressModes.Accumulator: A = (byte)data; break; // Absolute mode retrieves the byte at the indicated memory location case AddressModes.Absolute: memory[GetImmWord()] = (byte)data; break; case AddressModes.AbsoluteX: memory[GetImmWord() + X] = (byte)data; break; case AddressModes.AbsoluteY: memory[GetImmWord() + Y] = (byte)data; break; // Immediate mode uses the next byte in the instruction directly. case AddressModes.Immediate: throw new InvalidOperationException("Address mode " + mode.ToString() + " is not valid for this operation"); // Implied or Implicit are single byte instructions that do not use // the next bytes for the operand. case AddressModes.Implied: throw new InvalidOperationException("Address mode " + mode.ToString() + " is not valid for this operation"); // Indirect mode uses the absolute address to get another address. // The immediate word is a memory location from which to retrieve // the 16 bit operand. case AddressModes.Indirect: throw new InvalidOperationException("Address mode " + mode.ToString() + " is not valid for this operation"); // The indexed indirect modes uses the immediate byte rather than the // immediate word to get the memory location from which to retrieve // the 16 bit operand. This is a combination of ZeroPage indexed and Indirect. case AddressModes.XIndirect: memory[GetWordFromMemory((byte)(GetImmByte() + X))] = (byte)data; break; // The Indirect Indexed works a bit differently than above. // The Y register is added *after* the deferencing instead of before. case AddressModes.IndirectY: memory[GetWordFromMemory(GetImmByte()) + Y] = (byte)data; break; // Relative is used for branching, the immediate value is a // signed 8 bit value and used to offset the current PC. case AddressModes.Relative: throw new InvalidOperationException("Address mode " + mode.ToString() + " is not valid for this operation"); // Zero Page mode is a fast way of accessing the first 256 bytes of memory. // Best programming practice is to place your variables in 0x00-0xff. // Retrieve the byte at the indicated memory location. case AddressModes.ZeroPage: memory[GetImmByte()] = (byte)data; break; case AddressModes.ZeroPageX: memory[(GetImmByte() + X) & 0xff] = (byte)data; break; case AddressModes.ZeroPageY: memory[(GetImmByte() + Y) & 0xff] = (byte)data; break; case AddressModes.ZeroPage0: memory[GetWordFromMemory((GetImmByte()) & 0xff)] = (byte)data; break; // for this mode do the same thing as ZeroPage case AddressModes.BranchExt: memory[GetImmByte()] = (byte)data; break; default: break; } } private ushort GetWordFromMemory(int address) { return (ushort)((memory[address + 1] << 8 | memory[address]) & 0xffff); } private ushort GetImmWord() { return (ushort)((memory[PC + 2] << 8 | memory[PC + 1]) & 0xffff); } private byte GetImmByte() { return memory[PC + 1]; } private int SignExtend(int num) { if (num < 0x80) return num; else return (0xff << 8 | num) & 0xffff; } private void Push(byte data) { memory[(0x0100 | SP)] = data; SP--; } private void Push(ushort data) { // HI byte is in a higher address, LO byte is in the lower address memory[(0x0100 | SP)] = (byte)(data >> 8); memory[(0x0100 | (SP-1))] = (byte)(data & 0xff); SP -= 2; } private byte PopByte() { SP++; return memory[(0x0100 | SP)]; } private ushort PopWord() { // HI byte is in a higher address, LO byte is in the lower address SP += 2; ushort idx = (ushort)(0x0100 | SP); return (ushort)((memory[idx] << 8 | memory[idx-1]) & 0xffff); } private void ADC(byte oper) { ushort answer = (ushort)(A + oper); if (CF) answer++; CF = (answer > 0xff); ZF = ((answer & 0xff) == 0x00); NF = (answer & 0x80) == 0x80; //ushort temp = (ushort)(~(A ^ oper) & (A ^ answer) & 0x80); VF = (~(A ^ oper) & (A ^ answer) & 0x80) != 0x00; A = (byte)answer; } private int HexToBCD(byte oper) { // validate input is valid packed BCD if (oper > 0x99) throw new InvalidOperationException("Invalid BCD number: " + oper.ToString("X2")); if ((oper & 0x0f) > 0x09) throw new InvalidOperationException("Invalid BCD number: " + oper.ToString("X2")); return ((oper >> 4) * 10) + (oper & 0x0f); } private byte BCDToHex(int result) { if (result > 0xff) throw new InvalidOperationException("Invalid BCD to hex number: " + result.ToString()); if (result <= 9) return (byte)result; else return (byte)(((result / 10) << 4) + (result % 10)); } private void DoIRQ(ushort vector) { DoIRQ(vector, false); } private void DoIRQ(ushort vector, bool isBRK) { // Push the MSB of the PC Push((byte)(PC >> 8)); // Push the LSB of the PC Push((byte)(PC & 0xff)); // Push the status register int sr = 0x00; if (NF) sr = sr | 0x80; if (VF) sr = sr | 0x40; sr = sr | 0x20; // bit 5 is unused and always 1 if(isBRK) sr = sr | 0x10; // software interrupt (BRK) pushes B flag as 1 // hardware interrupt pushes B flag as 0 if (DF) sr = sr | 0x08; if (IF) sr = sr | 0x04; if (ZF) sr = sr | 0x02; if (CF) sr = sr | 0x01; Push((byte)sr); // set interrupt disable flag IF = true; // On 65C02, IRQ, NMI, and RESET also clear the D flag (but not on BRK) after pushing the status register. if (_cpuType == e6502Type.CMOS && !isBRK) DF = false; // load program counter with the interrupt vector PC = GetWordFromMemory(vector); } private void CheckBranch(bool flag, int oper) { if (flag) { // extra cycle on branch taken _extraCycles++; // extra cycle if branch destination is a different page than // the next instruction if ((PC & 0xff00) != ((PC + oper) & 0xff00)) _extraCycles++; PC += (ushort)oper; } } } } 

Commentaires

• Personne nest bienvenu vous à PPCG, je suppose que je ‘ vais saisir cette chance. Cest une excellente première réponse, et jespère vous voir plus souvent. Amusez-vous bien!
• Merci @StanStrum!Cétait un article SE il y a des années sur un émulateur 8086 qui ma intéressé à lémulation et à apprendre comment ces appareils fonctionnaient réellement. Cela ‘ a été très amusant. Outre ce qui précède, jai un émulateur 8080 complet et un émulateur 8086 qui est fait à environ 90%.
• Ce ‘ est génial, je ‘ Jai été intéressé par la création dun émulateur et / ou dun langage de programmation de niveau intermédiaire, mais je nai ‘ pas le temps, la patience ou lintellect pour le faire