Skip to content

Instantly share code, notes, and snippets.

@LongJohnCoder

LongJohnCoder/LICENSE

Forked from binji/LICENSE
Created June 5, 2021 13:40
Show Gist options
  • Save LongJohnCoder/2df13db938a2edf39e35ea93f31e279c to your computer and use it in GitHub Desktop.
Save LongJohnCoder/2df13db938a2edf39e35ea93f31e279c to your computer and use it in GitHub Desktop.
Download ZIP

Revisions

  1. @binji binji revised this gist Jun 3, 2021. 2 changed files with 35 additions and 0 deletions.
    17 changes: 17 additions & 0 deletions LICENSE
    View file
    Original file line number Diff line number Diff line change
    @@ -0,0 +1,17 @@
    Permission is hereby granted, free of charge, to any person obtaining a copy
    of this software and associated documentation files (the "Software"), to deal
    in the Software without restriction, including without limitation the rights
    to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
    copies of the Software, and to permit persons to whom the Software is
    furnished to do so, subject to the following conditions:

    The above copyright notice and this permission notice shall be included in
    all copies or substantial portions of the Software.

    THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
    IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
    FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
    AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
    LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
    OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
    THE SOFTWARE.
    18 changes: 18 additions & 0 deletions pokegb.cc
    View file
    Original file line number Diff line number Diff line change
    @@ -1,3 +1,10 @@
    /*
    * Copyright (C) 2021 Ben Smith
    *
    * This software may be modified and distributed under the terms
    * of the MIT license. See the LICENSE file for details.
    */

    #include <SDL2/SDL.h>
    #include <cstdint>
    #include <fcntl.h>
    @@ -81,13 +88,15 @@ uint8_t mem_access(uint16_t addr, uint8_t val, int write) {
    tick();
    switch (addr >> 12) {
    case 2: case 3:
    // Change ROM bank
    if (write)
    rom1 = rom0 + ((val ? val & 63 : 1) << 14);

    case 0: case 1:
    return rom0[addr];

    case 4: case 5:
    // Change RAM bank
    if (write && val <= 3)
    extrambank = extram + (val << 13);

    @@ -109,11 +118,14 @@ uint8_t mem_access(uint16_t addr, uint8_t val, int write) {
    case 15:
    if (addr >= 65024) {
    if (write) {
    // DMA transfer to OAM
    if (addr == 65350)
    for (int y = 160; --y >= 0;)
    io[y] = mem_access(val * 256 + y, 0, 0);
    io[addr & 511] = val;
    }

    // Read JOYP
    if (addr == 65280) {
    if (!(io[256] & 16))
    return ~(16 + key_state[SDL_SCANCODE_DOWN] * 8 +
    @@ -180,6 +192,7 @@ int main() {
    while (1) {
    prev_cycles = cycles;
    if (IME & IF & io[511]) {
    // Handle interrupt (VBlank only)
    IF = halt = IME = 0;
    cycles += 8;
    carry = 1;
    @@ -497,12 +510,14 @@ int main() {

    for (DIV += cycles - prev_cycles; prev_cycles++ != cycles;)
    if (LCDC & 128) {
    // Trigger VBlank interrupt
    if (++ppu_dot == 1 && LY == 144)
    IF |= 1;

    if (ppu_dot == 456) {
    if (LY < 144)
    for (tmp = 160; --tmp >= 0;) {
    // Draw background or window
    uint8_t is_window =
    LCDC & 32 && LY >= io[330] && tmp >= io[331] - 7,
    x_offset = is_window ? tmp - io[331] + 7 : tmp + io[323],
    @@ -521,6 +536,7 @@ int main() {
    color = (tile_data[1] >> x_offset) % 2 * 2 +
    (*tile_data >> x_offset) % 2;

    // Draw sprites
    if (LCDC & 2)
    for (uint8_t *sprite = io; sprite < io + 160; sprite += 4) {
    uint8_t sprite_x = tmp - sprite[1] + 8,
    @@ -542,12 +558,14 @@ int main() {
    }
    }

    // Draw pixel
    frame_buffer[LY * 160 + tmp] =
    palette[(io[327 + palette_index] >> (2 * color)) % 4 +
    palette_index * 4];
    }

    if (LY == 144) {
    // Render frame buffer
    void *pixels;
    SDL_LockTexture(texture, 0, &pixels, &tmp);
    for (tmp2 = 144; --tmp2 >= 0;)
  2. @binji binji revised this gist Jun 3, 2021. 1 changed file with 13 additions and 13 deletions.
    26 changes: 13 additions & 13 deletions pokegb.cc
    View file
    Original file line number Diff line number Diff line change
    @@ -153,9 +153,9 @@ uint16_t read16(uint16_t &addr = PC) {
    return read8(addr++) * 256 + tmp8;
    }

    void push(uint16_t addr) {
    write8(addr >> 8, --SP);
    write8(addr, --SP);
    void push(uint16_t val) {
    write8(val >> 8, --SP);
    write8(val, --SP);
    tick();
    }

    @@ -170,11 +170,11 @@ int main() {
    LCDC = 145;
    DIV = 44032;
    SDL_Init(SDL_INIT_VIDEO);
    SDL_Renderer *Sr = SDL_CreateRenderer(
    SDL_Renderer *renderer = SDL_CreateRenderer(
    SDL_CreateWindow("pokegb", 0, 0, 800, 720, SDL_WINDOW_SHOWN), -1,
    SDL_RENDERER_PRESENTVSYNC);
    SDL_Texture *St = SDL_CreateTexture(Sr, SDL_PIXELFORMAT_RGBA32,
    SDL_TEXTUREACCESS_STREAMING, 160, 144);
    SDL_Texture *texture = SDL_CreateTexture(
    renderer, SDL_PIXELFORMAT_RGBA32, SDL_TEXTUREACCESS_STREAMING, 160, 144);
    key_state = SDL_GetKeyboardState(0);

    while (1) {
    @@ -549,16 +549,16 @@ int main() {

    if (LY == 144) {
    void *pixels;
    SDL_LockTexture(St, 0, &pixels, &tmp);
    SDL_LockTexture(texture, 0, &pixels, &tmp);
    for (tmp2 = 144; --tmp2 >= 0;)
    memcpy((uint8_t *)pixels + tmp2 * tmp, frame_buffer + tmp2 * 160,
    640);
    SDL_UnlockTexture(St);
    SDL_RenderCopy(Sr, St, 0, 0);
    SDL_RenderPresent(Sr);
    SDL_Event e;
    while (SDL_PollEvent(&e))
    if (e.type == SDL_QUIT)
    SDL_UnlockTexture(texture);
    SDL_RenderCopy(renderer, texture, 0, 0);
    SDL_RenderPresent(renderer);
    SDL_Event event;
    while (SDL_PollEvent(&event))
    if (event.type == SDL_QUIT)
    return 0;
    }

  3. @binji binji revised this gist May 31, 2021. 1 changed file with 69 additions and 63 deletions.
    132 changes: 69 additions & 63 deletions pokegb.cc
    View file
    Original file line number Diff line number Diff line change
    @@ -116,9 +116,15 @@ uint8_t mem_access(uint16_t addr, uint8_t val, int write) {
    }
    if (addr == 65280) {
    if (!(io[256] & 16))
    return ~(16 + key_state[81] * 8 + key_state[82] * 4 + key_state[80] * 2 + key_state[79]);
    return ~(16 + key_state[SDL_SCANCODE_DOWN] * 8 +
    key_state[SDL_SCANCODE_UP] * 4 +
    key_state[SDL_SCANCODE_LEFT] * 2 +
    key_state[SDL_SCANCODE_RIGHT]);
    if (!(io[256] & 32))
    return ~(32 + key_state[40] * 8 + key_state[43] * 4 + key_state[29] * 2 + key_state[27]);
    return ~(32 + key_state[SDL_SCANCODE_RETURN] * 8 +
    key_state[SDL_SCANCODE_TAB] * 4 +
    key_state[SDL_SCANCODE_Z] * 2 +
    key_state[SDL_SCANCODE_X]);
    return 255;
    }
    return io[addr & 511];
    @@ -183,86 +189,86 @@ int main() {
    tick();
    else
    switch (opcode = read8_pc()) {
    case 0:
    case 0: // NOP
    break;

    OP4_NX16_REL(1)
    OP4_NX16_REL(1) // LD r16, u16
    *reg16_group1[opcode >> 4] = read16();
    break;

    OP4_NX16_REL(2)
    OP4_NX16_REL(2) // LD (r16), A
    write8(A, *reg16_group2[opcode >> 4]);
    HL += HL_add[opcode_rel / 16];
    break;

    OP4_NX16_REL(3)
    OP4_NX16_REL(3) // INC r16
    (*reg16_group1[opcode >> 4])++;
    tick();
    break;

    OP7_NX8_PTR(4)
    OP7_NX8_PTR(4) // INC r8
    tmp8 = ++(*ptr8);
    goto INC_FLAGS;

    case 52:
    case 52: // INC (HL)
    write8(tmp8 = read8() + 1);
    INC_FLAGS:
    set_flags(16, tmp8, 0, !(tmp8 & 15), 0);
    break;

    OP7_NX8_PTR(5)
    OP7_NX8_PTR(5) // DEC r8
    tmp8 = --(*ptr8);
    goto DEC_FLAGS;

    case 53:
    case 53: // DEC (HL)
    write8(tmp8 = read8() - 1);
    DEC_FLAGS:
    set_flags(16, tmp8, 1, tmp8 % 16 == 15, 0);
    break;

    OP7_NX8_PTR(6)
    OP7_NX8_PTR(6) // LD r8, u8
    *ptr8 = read8_pc();
    break;

    case 7:
    case 7: // RLCA
    A += A + (carry = A >> 7);
    goto CARRY_FLAG;

    OP4_NX16_REL(9)
    OP4_NX16_REL(9) // ADD HL, r16
    ptr16 = reg16_group1[opcode >> 4];
    set_flags(128, 1, 0, HL % 4096 + *ptr16 % 4096 > 4095, HL + *ptr16 > 65535);
    HL += *ptr16;
    tick();
    break;

    OP4_NX16_REL(10)
    OP4_NX16_REL(10) // LD A, (r16)
    A = read8(*reg16_group2[opcode >> 4]);
    HL += HL_add[opcode_rel / 16];
    break;

    OP4_NX16_REL(11)
    OP4_NX16_REL(11) // DEC r16
    (*reg16_group1[opcode >> 4])--;
    tick();
    break;

    case 15:
    case 15: // RRCA
    A = (carry = A & 1) * 128 + A / 2;
    goto CARRY_FLAG;

    case 23:
    case 23: // RLA
    carry = A >> 7;
    A += A + F / 16 % 2;
    CARRY_FLAG:
    set_flags(0, 1, 0, 0, carry);
    break;

    OP5_FLAG(32, 24)
    OP5_FLAG(32, 24) // JR i8 / JR <condition>, i8
    tmp8 = read8_pc();
    if (carry)
    PC += (int8_t)tmp8, tick();
    break;

    case 39:
    case 39: // DAA
    carry = tmp8 = 0;
    if (F & 32 || (!(F & 64) && A % 15 > 9))
    tmp8 = 6;
    @@ -271,49 +277,49 @@ int main() {
    set_flags(65, A += (F & 64) ? -tmp8 : tmp8, 0, 0, carry);
    break;

    case 47:
    case 47: // CPL
    A = ~A;
    set_flags(129, 1, 1, 1, 0);
    break;

    case 54:
    case 54: // LD (HL), u8
    write8(read8_pc());
    break;

    case 55: case 63:
    case 55: case 63: // SCF / CCF
    set_flags(128, 1, 0, 0, opcode == 55 ? 1 : !(F & 16));
    break;

    OP7_NX8_PTR(70)
    OP7_NX8_PTR(70) // LD r8, (HL)
    *ptr8 = read8();
    break;

    OP49_REL(64)
    OP49_REL(64) // LD r8, r8
    *reg8_group[opcode_rel / 8] = *reg8_group[opcode & 7];
    break;

    OP7_PTR(112)
    OP7_PTR(112) // LD (HL), r8
    write8(*ptr8);
    break;

    case 118:
    case 118: // HALT
    halt = 1;
    break;

    OP9_IMM_PTR(128)
    OP9_IMM_PTR(128) // ADD A, r8 / ADD A, (HL) / ADD A, u8
    neg = carry = 0;
    goto ALU;

    OP9_IMM_PTR(136)
    OP9_IMM_PTR(136) // ADC A, r8 / ADC A, (HL) / ADC A, u8
    neg = 0;
    carry = F / 16 % 2;
    goto ALU;

    OP9_IMM_PTR(144)
    OP9_IMM_PTR(144) // SUB A, r8 / SUB A, (HL) / SUB A, u8
    carry = 1;
    goto SUBTRACT;

    OP9_IMM_PTR(152)
    OP9_IMM_PTR(152) // SBC A, r8 / SBC A, (HL) / SBC A, u8
    carry = !(F / 16 % 2);
    SUBTRACT:
    neg = 1;
    @@ -325,165 +331,165 @@ int main() {
    A = tmp8;
    break;

    OP9_IMM_PTR(160)
    OP9_IMM_PTR(160) // AND A, r8 / AND A, (HL) / AND A, u8
    set_flags(0, A &= operand, 0, 1, 0);
    break;

    OP9_IMM_PTR(168)
    OP9_IMM_PTR(168) // XOR A, r8 / XOR A, (HL) / XOR A, u8
    set_flags(0, A ^= operand, 0, 0, 0);
    break;

    OP9_IMM_PTR(176)
    OP9_IMM_PTR(176) // OR A, r8 / OR A, (HL) / OR A, u8
    set_flags(0, A |= operand, 0, 0, 0);
    break;

    OP9_IMM_PTR(184)
    OP9_IMM_PTR(184) // CP A, r8 / CP A, (HL) / CP A, u8
    set_flags(0, A != operand, 1, A % 16 < operand % 16, A < operand);
    break;

    case 217:
    case 217: // RETI
    carry = IME = 1;
    goto RET;

    OP5_FLAG(192, 201)
    OP5_FLAG(192, 201) // RET / RET <condition>
    RET:
    tick();
    if (carry)
    PC = read16(SP);
    break;

    OP4_NX16_REL(193)
    OP4_NX16_REL(193) // POP r16
    reg16[opcode_rel >> 4] = read16(SP);
    break;

    OP5_FLAG(194, 195)
    OP5_FLAG(194, 195) // JP u16 / JP <condition>, u16
    temp16 = read16();
    if (carry)
    PC = temp16, tick();
    break;

    OP5_FLAG(196, 205)
    OP5_FLAG(196, 205) // CALL u16 / CALL <condition>, u16
    temp16 = read16();
    CALL:
    if (carry)
    push(PC), PC = temp16;
    break;

    OP4_NX16_REL(197)
    OP4_NX16_REL(197) // PUSH r16
    push(reg16[opcode_rel >> 4]);
    break;

    case 203:
    switch (opcode = read8_pc()) {
    OP7_PTR(0)
    OP7_PTR(0) // RLC r8
    *ptr8 += *ptr8 + (carry = *ptr8 / 128 % 2);
    goto CARRY_ZERO_FLAGS_PTR;

    OP7_PTR(8)
    OP7_PTR(8) // RRC r8
    *ptr8 = (carry = *ptr8 & 1) * 128 + *ptr8 / 2;
    goto CARRY_ZERO_FLAGS_PTR;

    case 14:
    case 14: // RRC (HL)
    tmp8 = read8();
    carry = tmp8 & 1;
    write8(tmp8 = carry * 128 + tmp8 / 2);
    goto CARRY_ZERO_FLAGS_U;

    OP7_PTR(16)
    OP7_PTR(16) // RL r8
    carry = *ptr8 >> 7;
    *ptr8 = *ptr8 * 2 + F / 16 % 2;
    goto CARRY_ZERO_FLAGS_PTR;

    OP7_PTR(24)
    OP7_PTR(24) // RR r8
    carry = *ptr8 & 1;
    *ptr8 = *ptr8 / 2 + ((F * 8) & 128);
    goto CARRY_ZERO_FLAGS_PTR;

    OP7_PTR(32)
    OP7_PTR(32) // SLA r8
    carry = *ptr8 / 128 % 2;
    *ptr8 *= 2;
    goto CARRY_ZERO_FLAGS_PTR;

    OP7_PTR(40)
    OP7_PTR(40) // SRA r8
    carry = *ptr8 & 1;
    *ptr8 = (int8_t)*ptr8 >> 1;
    goto CARRY_ZERO_FLAGS_PTR;

    OP7_PTR(48)
    OP7_PTR(48) // SWAP r8
    set_flags(0, *ptr8 = *ptr8 * 16 + *ptr8 / 16, 0, 0, 0);
    break;

    case 54:
    case 54: // SWAP (HL)
    tmp8 = read8();
    carry = 0;
    write8(tmp8 = tmp8 * 16 + tmp8 / 16);
    CARRY_ZERO_FLAGS_U:
    set_flags(0, tmp8, 0, 0, carry);
    break;

    OP7_PTR(56)
    OP7_PTR(56) // SRL r8
    carry = *ptr8 & 1;
    *ptr8 /= 2;
    CARRY_ZERO_FLAGS_PTR:
    set_flags(0, *ptr8, 0, 0, carry);
    break;

    OP8_NX8_REL(70)
    OP8_NX8_REL(70) // BIT bit, r8
    tmp8 = read8();
    goto BIT_FLAGS;

    OP56_PTR_REL(64)
    OP56_PTR_REL(64) // BIT bit, (HL)
    tmp8 = *ptr8;
    BIT_FLAGS:
    set_flags(16, tmp8 & (1 << opcode_rel / 8), 0, 1, 0);
    break;

    OP8_NX8_REL(134)
    OP8_NX8_REL(134) // RES bit, (HL)
    write8(read8() & ~(1 << opcode_rel / 8));
    break;

    OP56_PTR_REL(128)
    OP56_PTR_REL(128) // RES bit, r8
    *ptr8 &= ~(1 << opcode_rel / 8);
    break;

    OP8_NX8_REL(198)
    OP8_NX8_REL(198) // SET bit, (HL)
    write8(read8() | (1 << opcode_rel / 8));
    break;

    OP56_PTR_REL(192)
    OP56_PTR_REL(192) // SET bit, r8
    *ptr8 |= 1 << opcode_rel / 8;
    break;
    }
    break;

    case 224: case 226:
    case 224: case 226: // LD (FF00 + u8), A / LD (FF00 + C), A
    write8(A, 65280 + (opcode == 224 ? read8_pc() : *reg8));
    break;

    case 233:
    case 233: // JP HL
    PC = HL;
    break;

    case 234:
    case 234: // LD (u16), A
    write8(A, read16());
    break;

    case 240: case 250:
    case 240: case 250: // LD A, (FF00 + u8) / LD A, (u16)
    A = read8(opcode == 240 ? 65280 | read8_pc() : read16());
    break;

    case 243: case 251:
    case 243: case 251: // DI / EI
    IME = opcode != 243;
    break;

    case 248:
    case 248: // LD HL, SP + i8
    tmp8 = read8_pc();
    set_flags(0, 1, 0, (uint8_t)SP + tmp8 > 255, SP % 16 + tmp8 % 16 > 15);
    HL = SP + (int8_t)tmp8;
    tick();
    break;

    case 249:
    case 249: // LD SP, HL
    SP = HL;
    tick();
    break;
  4. @binji binji revised this gist May 29, 2021. 1 changed file with 483 additions and 788 deletions.
    1,271 changes: 483 additions & 788 deletions pokegb.cc
    View file
    Original file line number Diff line number Diff line change
    @@ -1,870 +1,565 @@
    using u8 = uint8_t;
    using s8 = int8_t;
    using u16 = uint16_t;
    using u32 = uint32_t;
    using u64 = uint64_t;

    u8 op, u, v, fc, *rom0, *rom1, io[512], vram[8192], wram[16384], *eram, *eram1,
    R[] = {19, 0, 216, 0, 77, 1, 176, 1, 254, 255}, &F = R[6], &A = R[7],
    *R8[] = {&R[1], &R[0], &R[3], &R[2], &R[5], &R[4], &F, &A}, &IF = io[271],
    &LCDC = io[320], &LY = io[324], &WY = io[330], &WX = io[331], &IE = io[511],
    IME = 0, halt = 0, FM[] = {128, 128, 16, 16}, FE[] = {0, 128, 0, 16}, *p;
    u8 const *Sk;
    u16 PC = 256, U, *RR = (u16 *)R, &HL = RR[2], &SP = RR[4],
    *RR2[] = {&RR[0], &RR[1], &HL, &SP}, *RR3[] = {&RR[0], &RR[1], &HL, &HL},
    &DIV = (u16 &)io[259], dot = 32, *P;
    u32 fb[23040], pal[] = {0xffffffff, 0xffaaaaaa, 0xff555555, 0xff000000};
    u64 bcy, cy = 0;
    int HA[] = {0, 0, 1, -1};

    void T() { cy += 4; }
    u8 mem(u16 a, u8 x, int w) {
    T();
    switch (a >> 12) {
    default:
    case 2 ... 3:
    if (w)
    rom1 = rom0 + ((x ? (x & 63) : 1) << 14);
    case 0 ... 1:
    return rom0[a];
    case 4 ... 5:
    if (w && x <= 3)
    eram1 = eram + (x << 13);
    case 6 ... 7:
    return rom1[a & 16383];
    case 8 ... 9:
    a &= 8191;
    if (w)
    vram[a] = x;
    return vram[a];
    case 10 ... 11:
    a &= 8191;
    if (w)
    eram1[a] = x;
    return eram1[a];
    case 15:
    if (a >= 65024) {
    if (w) {
    if (a == 65350)
    for (int i = 160; --i >= 0;)
    io[i] = mem(x * 256 + i, 0, 0);
    io[a & 511] = x;
    }
    if (a == 65280) {
    if (!(io[256] & 16))
    return ~(16 + Sk[81] * 8 + (Sk[82] * 4) + Sk[80] * 2 + Sk[79]);
    if (!(io[256] & 32))
    return ~(32 + Sk[40] * 8 + Sk[43] * 4 + Sk[29] * 2 + Sk[27]);
    return 255;
    #include <SDL2/SDL.h>
    #include <cstdint>
    #include <fcntl.h>
    #include <sys/mman.h>
    #include <unistd.h>

    #define OP4_NX8(_) case _: case _ + 8: case _ + 16: case _ + 24:

    #define OP4_NX16_REL(_) \
    case _: \
    case _ + 16: \
    case _ + 32: \
    case _ + 48: \
    opcode_rel = opcode - _;

    #define OP5_FLAG(_, always) \
    OP4_NX8(_) \
    case always: \
    carry = (opcode == always) || \
    (F & F_mask[(opcode - _) / 8]) == F_equals[(opcode - _) / 8];

    #define OP8_NX8_REL(_) OP4_NX8(_) OP4_NX8(_ + 32) opcode_rel = opcode - _;

    #define OP7_PTR(_) \
    case _: \
    case _ + 1: \
    case _ + 2: \
    case _ + 3: \
    case _ + 4: \
    case _ + 5: \
    case _ + 7: \
    ptr8 = reg8_group[opcode & 7];

    #define OP7_NX8(_) OP4_NX8(_) case _ + 32 : case _ + 40 : case _ + 56:

    #define OP7_NX8_PTR(_) OP7_NX8(_) ptr8 = reg8_group[(opcode - _) / 8];

    #define OP49_REL(_) \
    OP7_NX8(_) \
    OP7_NX8(_ + 1) \
    OP7_NX8(_ + 2) \
    OP7_NX8(_ + 3) \
    OP7_NX8(_ + 4) \
    OP7_NX8(_ + 5) \
    OP7_NX8(_ + 7) \
    opcode_rel = opcode - _;

    #define OP56_PTR_REL(_) OP49_REL(_) OP7_PTR(_ + 48) opcode_rel = opcode - _;

    #define OP9_IMM_PTR(_) \
    case _ + 6: \
    case _ + 70: \
    OP7_PTR(_) \
    operand = (opcode == _ + 6) ? read8() \
    : (opcode == _ + 70) ? read8_pc() \
    : *ptr8;

    uint8_t opcode, opcode_rel, tmp8, operand, carry, neg, *rom0, *rom1, io[512], video_ram[8192],
    work_ram[16384], *extram, *extrambank,
    reg8[] = {19, 0, 216, 0, 77, 1, 176, 1, 254, 255}, &F = reg8[6],
    &A = reg8[7], *reg8_group[] = {reg8 + 1, reg8, reg8 + 3, reg8 + 2,
    reg8 + 5, reg8 + 4, &F, &A},
    &IF = io[271], &LCDC = io[320], &LY = io[324], IME, halt, *ptr8;

    uint8_t const *key_state;

    uint16_t PC = 256, temp16, *reg16 = (uint16_t *)reg8, &HL = reg16[2],
    &SP = reg16[4], &DIV = (uint16_t &)io[259], ppu_dot = 32, *ptr16,
    *reg16_group1[] = {reg16, reg16 + 1, &HL, &SP},
    *reg16_group2[] = {reg16, reg16 + 1, &HL, &HL}, prev_cycles, cycles;

    int tmp, tmp2,
    HL_add[] = {0, 0, 1, -1}, F_mask[] = {128, 128, 16, 16},
    F_equals[] = {0, 128, 0, 16}, frame_buffer[23040],
    palette[] = {-1, -23197, -65536, -16777216, -1, -8092417,
    -12961132, -16777216, -1, -23197, -65536, -16777216};

    void tick() { cycles += 4; }

    uint8_t mem_access(uint16_t addr, uint8_t val, int write) {
    tick();
    switch (addr >> 12) {
    case 2: case 3:
    if (write)
    rom1 = rom0 + ((val ? val & 63 : 1) << 14);

    case 0: case 1:
    return rom0[addr];

    case 4: case 5:
    if (write && val <= 3)
    extrambank = extram + (val << 13);

    case 6: case 7:
    return rom1[addr & 16383];

    case 8: case 9:
    addr &= 8191;
    if (write)
    video_ram[addr] = val;
    return video_ram[addr];

    case 10: case 11:
    addr &= 8191;
    if (write)
    extrambank[addr] = val;
    return extrambank[addr];

    case 15:
    if (addr >= 65024) {
    if (write) {
    if (addr == 65350)
    for (int y = 160; --y >= 0;)
    io[y] = mem_access(val * 256 + y, 0, 0);
    io[addr & 511] = val;
    }
    if (addr == 65280) {
    if (!(io[256] & 16))
    return ~(16 + key_state[81] * 8 + key_state[82] * 4 + key_state[80] * 2 + key_state[79]);
    if (!(io[256] & 32))
    return ~(32 + key_state[40] * 8 + key_state[43] * 4 + key_state[29] * 2 + key_state[27]);
    return 255;
    }
    return io[addr & 511];
    }
    return io[a & 511];
    }
    case 12 ... 14:
    a &= 16383;
    if (w)
    wram[a] = x;
    return wram[a];

    case 12 ... 14:
    addr &= 16383;
    if (write)
    work_ram[addr] = val;
    return work_ram[addr];
    }
    }
    u8 r(u16 a) { return mem(a, 0, 0); }
    void w(u16 a, u8 x) { mem(a, x, 1); }
    void FS(u8 M, int Z, int N, int H, int C) {
    F = (F & M) | (!Z * 128 + N * 64 + H * 32 + C * 16);

    uint8_t read8(uint16_t addr = HL) { return mem_access(addr, 0, 0); }

    void write8(uint8_t val, uint16_t addr = HL) { mem_access(addr, val, 1); }

    void set_flags(uint8_t mask, int Z, int N, int H, int C) {
    F = (F & mask) | (!Z * 128 + N * 64 + H * 32 + C * 16);
    }
    u8 r8() { return r(PC++); }
    u16 r16(u16 &R = PC) {
    u = r(R++);
    return r(R++) * 256 + u;

    uint8_t read8_pc() { return read8(PC++); }

    uint16_t read16(uint16_t &addr = PC) {
    tmp8 = read8(addr++);
    return read8(addr++) * 256 + tmp8;
    }
    void push(u16 x) {
    w(--SP, x >> 8);
    w(--SP, x);
    T();

    void push(uint16_t addr) {
    write8(addr >> 8, --SP);
    write8(addr, --SP);
    tick();
    }

    int main() {
    rom1 =
    (rom0 = (u8 *)mmap(0, 1048576, 0x1, 0x01, open("rom.gb", 00), 0)) + 32768;
    int sav = open("rom.sav", 0100 | 02, 0666);
    ftruncate(sav, 32768);
    eram1 = eram = (u8 *)mmap(0, 32768, 0x1 | 0x2, 0x01, sav, 0);
    rom1 = (rom0 = (uint8_t *)mmap(0, 1048576, PROT_READ, MAP_SHARED,
    open("rom.gb", O_RDONLY), 0)) +
    32768;
    tmp = open("rom.sav", O_CREAT|O_RDWR, 0666);
    ftruncate(tmp, 32768);
    extrambank = extram =
    (uint8_t *)mmap(0, 32768, PROT_READ | PROT_WRITE, MAP_SHARED, tmp, 0);
    LCDC = 145;
    DIV = 44032;
    SDL_Init((0x00000001u | 0x00000010u | 0x00000020u | 0x00004000u |
    0x00000200u | 0x00001000u | 0x00002000u | 0x00008000u));
    SDL_Window *Sw =
    SDL_CreateWindow("pokegb", 100, 100, 160 * 5, 144 * 5, SDL_WINDOW_SHOWN);
    SDL_Renderer *Sr = SDL_CreateRenderer(Sw, -1, 0);
    SDL_Init(SDL_INIT_VIDEO);
    SDL_Renderer *Sr = SDL_CreateRenderer(
    SDL_CreateWindow("pokegb", 0, 0, 800, 720, SDL_WINDOW_SHOWN), -1,
    SDL_RENDERER_PRESENTVSYNC);
    SDL_Texture *St = SDL_CreateTexture(Sr, SDL_PIXELFORMAT_RGBA32,
    SDL_TEXTUREACCESS_STREAMING, 160, 144);
    Sk = SDL_GetKeyboardState(0);
    key_state = SDL_GetKeyboardState(0);

    while (1) {
    bcy = cy;
    if (IME & IF & IE) {
    prev_cycles = cycles;
    if (IME & IF & io[511]) {
    IF = halt = IME = 0;
    cy += 8;
    fc = 1;
    U = 64;
    goto CALLU;
    cycles += 8;
    carry = 1;
    temp16 = 64;
    goto CALL;
    } else if (halt)
    T();
    tick();
    else
    switch ((op = r8())) {
    switch (opcode = read8_pc()) {
    case 0:
    break;
    case 1:
    case 1 + 16:
    case 1 + 32:
    case 1 + 48:
    *RR2[op >> 4] = r16();
    break;
    case 2:
    case 2 + 16:
    case 2 + 32:
    case 2 + 48:
    w(*RR3[op >> 4], A);
    HL += HA[(op - 2) / 16];
    break;
    case 3:
    case 3 + 16:
    case 3 + 32:
    case 3 + 48:
    (*RR2[op >> 4])++;
    T();
    break;
    case 4:
    case 4 + 8:
    case 4 + 16:
    case 4 + 24:
    case 4 + 32:
    case 4 + 40:
    case 4 + 56:
    p = R8[(op - 4) / 8];
    ++(*p);
    FS(16, *p, 0, !(*p & 15), 0);
    break;
    case 5:
    case 5 + 8:
    case 5 + 16:
    case 5 + 24:
    case 5 + 32:
    case 5 + 40:
    case 5 + 56:
    p = R8[(op - 5) / 8];
    --(*p);
    FS(16, *p, 1, (*p & 15) == 15, 0);
    break;
    case 6:
    case 6 + 8:
    case 6 + 16:
    case 6 + 24:
    case 6 + 32:
    case 6 + 40:
    case 6 + 56:
    p = R8[(op - 6) / 8];
    *p = r8();

    OP4_NX16_REL(1)
    *reg16_group1[opcode >> 4] = read16();
    break;

    OP4_NX16_REL(2)
    write8(A, *reg16_group2[opcode >> 4]);
    HL += HL_add[opcode_rel / 16];
    break;

    OP4_NX16_REL(3)
    (*reg16_group1[opcode >> 4])++;
    tick();
    break;

    OP7_NX8_PTR(4)
    tmp8 = ++(*ptr8);
    goto INC_FLAGS;

    case 52:
    write8(tmp8 = read8() + 1);
    INC_FLAGS:
    set_flags(16, tmp8, 0, !(tmp8 & 15), 0);
    break;

    OP7_NX8_PTR(5)
    tmp8 = --(*ptr8);
    goto DEC_FLAGS;

    case 53:
    write8(tmp8 = read8() - 1);
    DEC_FLAGS:
    set_flags(16, tmp8, 1, tmp8 % 16 == 15, 0);
    break;

    OP7_NX8_PTR(6)
    *ptr8 = read8_pc();
    break;

    case 7:
    A += A + (fc = A >> 7);
    FS(0, 1, 0, 0, fc);
    break;
    case 9:
    case 9 + 16:
    case 9 + 32:
    case 9 + 48:
    P = RR2[op >> 4];
    FS(128, 1, 0, HL % 4096 + *P % 4096 > 4095, HL + *P > 65535);
    HL += *P;
    T();
    break;
    case 10:
    case 10 + 16:
    case 10 + 32:
    case 10 + 48:
    A = r(*RR3[op >> 4]);
    HL += HA[(op - 10) / 16];
    break;
    case 11:
    case 11 + 16:
    case 11 + 32:
    case 11 + 48:
    (*RR2[op >> 4])--;
    T();
    A += A + (carry = A >> 7);
    goto CARRY_FLAG;

    OP4_NX16_REL(9)
    ptr16 = reg16_group1[opcode >> 4];
    set_flags(128, 1, 0, HL % 4096 + *ptr16 % 4096 > 4095, HL + *ptr16 > 65535);
    HL += *ptr16;
    tick();
    break;
    case 15:
    A = (fc = A & 1) * 128 + A / 2;
    FS(0, 1, 0, 0, fc);

    OP4_NX16_REL(10)
    A = read8(*reg16_group2[opcode >> 4]);
    HL += HL_add[opcode_rel / 16];
    break;

    OP4_NX16_REL(11)
    (*reg16_group1[opcode >> 4])--;
    tick();
    break;

    case 15:
    A = (carry = A & 1) * 128 + A / 2;
    goto CARRY_FLAG;

    case 23:
    fc = A >> 7;
    A += A + (F >> 4) % 2;
    FS(0, 1, 0, 0, fc);
    break;
    case 32:
    case 32 + 8:
    case 32 + 16:
    case 32 + 24:
    case 24:
    fc = (op == 24) || (F & FM[(op - 32) / 8]) == FE[(op - 32) / 8];
    u = r8();
    if (fc)
    PC += (s8)u, T();
    carry = A >> 7;
    A += A + F / 16 % 2;
    CARRY_FLAG:
    set_flags(0, 1, 0, 0, carry);
    break;

    OP5_FLAG(32, 24)
    tmp8 = read8_pc();
    if (carry)
    PC += (int8_t)tmp8, tick();
    break;

    case 39:
    fc = u = 0;
    carry = tmp8 = 0;
    if (F & 32 || (!(F & 64) && A % 15 > 9))
    u = 6;
    tmp8 = 6;
    if (F & 16 || (!(F & 64) && A > 153))
    u |= 96, fc = 1;
    FS(65, A += (F & 64) ? -u : u, 0, 0, fc);
    tmp8 |= 96, carry = 1;
    set_flags(65, A += (F & 64) ? -tmp8 : tmp8, 0, 0, carry);
    break;

    case 47:
    A = ~A;
    FS(129, 1, 1, 1, 0);
    set_flags(129, 1, 1, 1, 0);
    break;
    case 52:
    u = r(HL) + 1;
    FS(16, u, 0, !(u & 15), 0);
    w(HL, u);

    case 54:
    write8(read8_pc());
    break;
    case 53:
    u = r(HL) - 1;
    FS(16, u, 1, u % 16 == 15, 0);
    w(HL, u);

    case 55: case 63:
    set_flags(128, 1, 0, 0, opcode == 55 ? 1 : !(F & 16));
    break;
    case 54:
    w(HL, r8());
    break;
    case 55:
    FS(128, 1, 0, 0, 1);
    break;
    case 63:
    FS(128, 1, 0, 0, !(F & 16));
    break;
    case 70:
    case 70 + 8:
    case 70 + 16:
    case 70 + 24:
    case 70 + 32:
    case 70 + 40:
    case 70 + 56:
    p = R8[(op - 70) / 8];
    *p = r(HL);
    break;
    case 64:
    case 64 + 8:
    case 64 + 16:
    case 64 + 24:
    case 64 + 32:
    case 64 + 40:
    case 64 + 56:
    case 64 + 1:
    case 64 + 1 + 8:
    case 64 + 1 + 16:
    case 64 + 1 + 24:
    case 64 + 1 + 32:
    case 64 + 1 + 40:
    case 64 + 1 + 56:
    case 64 + 2:
    case 64 + 2 + 8:
    case 64 + 2 + 16:
    case 64 + 2 + 24:
    case 64 + 2 + 32:
    case 64 + 2 + 40:
    case 64 + 2 + 56:
    case 64 + 3:
    case 64 + 3 + 8:
    case 64 + 3 + 16:
    case 64 + 3 + 24:
    case 64 + 3 + 32:
    case 64 + 3 + 40:
    case 64 + 3 + 56:
    case 64 + 4:
    case 64 + 4 + 8:
    case 64 + 4 + 16:
    case 64 + 4 + 24:
    case 64 + 4 + 32:
    case 64 + 4 + 40:
    case 64 + 4 + 56:
    case 64 + 5:
    case 64 + 5 + 8:
    case 64 + 5 + 16:
    case 64 + 5 + 24:
    case 64 + 5 + 32:
    case 64 + 5 + 40:
    case 64 + 5 + 56:
    case 64 + 7:
    case 64 + 7 + 8:
    case 64 + 7 + 16:
    case 64 + 7 + 24:
    case 64 + 7 + 32:
    case 64 + 7 + 40:
    case 64 + 7 + 56:
    *R8[(op - 64) / 8] = *R8[op & 7];
    break;
    case 112:
    case 112 + 1:
    case 112 + 2:
    case 112 + 3:
    case 112 + 4:
    case 112 + 5:
    case 112 + 7:
    p = R8[op & 7];
    w(HL, *p);

    OP7_NX8_PTR(70)
    *ptr8 = read8();
    break;

    OP49_REL(64)
    *reg8_group[opcode_rel / 8] = *reg8_group[opcode & 7];
    break;

    OP7_PTR(112)
    write8(*ptr8);
    break;

    case 118:
    halt = 1;
    break;
    case 128 + 6:
    case 128 + 70:
    case 128:
    case 128 + 1:
    case 128 + 2:
    case 128 + 3:
    case 128 + 4:
    case 128 + 5:
    case 128 + 7:
    p = R8[op & 7];
    v = (op == 128 + 6) ? r(HL) : (op == 128 + 70) ? r8() : *p;
    fc = 0;
    goto ADD;
    case 136 + 6:
    case 136 + 70:
    case 136:
    case 136 + 1:
    case 136 + 2:
    case 136 + 3:
    case 136 + 4:
    case 136 + 5:
    case 136 + 7:
    p = R8[op & 7];
    v = (op == 136 + 6) ? r(HL) : (op == 136 + 70) ? r8() : *p;
    fc = (F >> 4) & 1;
    ADD:
    FS(0, u = A + v + fc, 0, A % 16 + v % 16 + fc > 15, A + v + fc > 255);
    A = u;
    break;
    case 144 + 6:
    case 144 + 70:
    case 144:
    case 144 + 1:
    case 144 + 2:
    case 144 + 3:
    case 144 + 4:
    case 144 + 5:
    case 144 + 7:
    p = R8[op & 7];
    v = (op == 144 + 6) ? r(HL) : (op == 144 + 70) ? r8() : *p;
    fc = 0;
    goto SUB;
    case 152 + 6:
    case 152 + 70:
    case 152:
    case 152 + 1:
    case 152 + 2:
    case 152 + 3:
    case 152 + 4:
    case 152 + 5:
    case 152 + 7:
    p = R8[op & 7];
    v = (op == 152 + 6) ? r(HL) : (op == 152 + 70) ? r8() : *p;
    fc = (F >> 4) & 1;
    SUB:
    FS(0, u = A - v - fc, 1, A % 16 < v % 16 + fc, A < v + fc);
    A = u;
    break;
    case 160 + 6:
    case 160 + 70:
    case 160:
    case 160 + 1:
    case 160 + 2:
    case 160 + 3:
    case 160 + 4:
    case 160 + 5:
    case 160 + 7:
    p = R8[op & 7];
    v = (op == 160 + 6) ? r(HL) : (op == 160 + 70) ? r8() : *p;
    FS(0, A &= v, 0, 1, 0);
    break;
    case 168 + 6:
    case 168 + 70:
    case 168:
    case 168 + 1:
    case 168 + 2:
    case 168 + 3:
    case 168 + 4:
    case 168 + 5:
    case 168 + 7:
    p = R8[op & 7];
    v = (op == 168 + 6) ? r(HL) : (op == 168 + 70) ? r8() : *p;
    FS(0, A ^= v, 0, 0, 0);
    break;
    case 176 + 6:
    case 176 + 70:
    case 176:
    case 176 + 1:
    case 176 + 2:
    case 176 + 3:
    case 176 + 4:
    case 176 + 5:
    case 176 + 7:
    p = R8[op & 7];
    v = (op == 176 + 6) ? r(HL) : (op == 176 + 70) ? r8() : *p;
    FS(0, A |= v, 0, 0, 0);
    break;
    case 184 + 6:
    case 184 + 70:
    case 184:
    case 184 + 1:
    case 184 + 2:
    case 184 + 3:
    case 184 + 4:
    case 184 + 5:
    case 184 + 7:
    p = R8[op & 7];
    v = (op == 184 + 6) ? r(HL) : (op == 184 + 70) ? r8() : *p;
    FS(0, A != v, 1, A % 16 < v % 16, A < v);

    OP9_IMM_PTR(128)
    neg = carry = 0;
    goto ALU;

    OP9_IMM_PTR(136)
    neg = 0;
    carry = F / 16 % 2;
    goto ALU;

    OP9_IMM_PTR(144)
    carry = 1;
    goto SUBTRACT;

    OP9_IMM_PTR(152)
    carry = !(F / 16 % 2);
    SUBTRACT:
    neg = 1;
    operand = ~operand;
    ALU:
    set_flags(0, tmp8 = A + operand + carry, neg,
    (A % 16 + operand % 16 + carry > 15) ^ neg,
    (A + operand + carry > 255) ^ neg);
    A = tmp8;
    break;

    OP9_IMM_PTR(160)
    set_flags(0, A &= operand, 0, 1, 0);
    break;

    OP9_IMM_PTR(168)
    set_flags(0, A ^= operand, 0, 0, 0);
    break;

    OP9_IMM_PTR(176)
    set_flags(0, A |= operand, 0, 0, 0);
    break;

    OP9_IMM_PTR(184)
    set_flags(0, A != operand, 1, A % 16 < operand % 16, A < operand);
    break;

    case 217:
    fc = IME = 1;
    carry = IME = 1;
    goto RET;
    case 192:
    case 192 + 8:
    case 192 + 16:
    case 192 + 24:
    case 201:
    fc = (op == 201) || (F & FM[(op - 192) / 8]) == FE[(op - 192) / 8];

    OP5_FLAG(192, 201)
    RET:
    T();
    if (fc)
    PC = r16(SP);
    break;
    case 193:
    case 193 + 16:
    case 193 + 32:
    case 193 + 48:
    RR[(op - 193) >> 4] = r16(SP);
    break;
    case 194:
    case 194 + 8:
    case 194 + 16:
    case 194 + 24:
    case 195:
    fc = (op == 195) || (F & FM[(op - 194) / 8]) == FE[(op - 194) / 8];
    U = r16();
    if (fc)
    PC = U, T();
    break;
    case 196:
    case 196 + 8:
    case 196 + 16:
    case 196 + 24:
    case 205:
    fc = (op == 205) || (F & FM[(op - 196) / 8]) == FE[(op - 196) / 8];
    U = r16();
    CALLU:
    if (fc)
    push(PC), PC = U;
    break;
    case 197:
    case 197 + 16:
    case 197 + 32:
    case 197 + 48:
    push(RR[(op - 197) >> 4]);
    tick();
    if (carry)
    PC = read16(SP);
    break;

    OP4_NX16_REL(193)
    reg16[opcode_rel >> 4] = read16(SP);
    break;

    OP5_FLAG(194, 195)
    temp16 = read16();
    if (carry)
    PC = temp16, tick();
    break;

    OP5_FLAG(196, 205)
    temp16 = read16();
    CALL:
    if (carry)
    push(PC), PC = temp16;
    break;

    OP4_NX16_REL(197)
    push(reg16[opcode_rel >> 4]);
    break;

    case 203:
    switch ((op = r8())) {
    case 7:
    A += A + (fc = (A >> 7) & 1);
    FS(0, A, 0, 0, fc);
    break;
    case 8:
    case 8 + 1:
    case 8 + 2:
    case 8 + 3:
    case 8 + 4:
    case 8 + 5:
    case 8 + 7:
    p = R8[op & 7];
    *p = (fc = *p & 1) * 128 + *p / 2;
    FS(0, *p, 0, 0, fc);
    break;
    case 14:
    u = r(HL);
    fc = u & 1;
    u = fc * 128 + u / 2;
    FS(0, u, 0, 0, fc);
    w(HL, u);
    break;
    case 16:
    case 16 + 1:
    case 16 + 2:
    case 16 + 3:
    case 16 + 4:
    case 16 + 5:
    case 16 + 7:
    p = R8[op & 7];
    fc = *p >> 7;
    FS(0, *p = *p * 2 + (F >> 4) % 2, 0, 0, fc);
    break;
    case 24:
    case 24 + 1:
    case 24 + 2:
    case 24 + 3:
    case 24 + 4:
    case 24 + 5:
    case 24 + 7:
    p = R8[op & 7];
    fc = *p & 1;
    FS(0, *p = *p / 2 + ((F * 8) & 128), 0, 0, fc);
    break;
    case 32:
    case 32 + 1:
    case 32 + 2:
    case 32 + 3:
    case 32 + 4:
    case 32 + 5:
    case 32 + 7:
    p = R8[op & 7];
    fc = (*p >> 7) & 1;
    FS(0, *p *= 2, 0, 0, fc);
    break;
    case 40:
    case 40 + 1:
    case 40 + 2:
    case 40 + 3:
    case 40 + 4:
    case 40 + 5:
    case 40 + 7:
    p = R8[op & 7];
    fc = *p & 1;
    FS(0, *p = (s8)*p >> 1, 0, 0, fc);
    break;
    case 48:
    case 48 + 1:
    case 48 + 2:
    case 48 + 3:
    case 48 + 4:
    case 48 + 5:
    case 48 + 7:
    p = R8[op & 7];
    FS(0, *p = *p * 16 + *p / 16, 0, 0, 0);
    break;
    case 54:
    u = r(HL);
    FS(0, u = u * 16 + u / 16, 0, 0, 0);
    w(HL, u);
    break;
    case 56:
    case 56 + 1:
    case 56 + 2:
    case 56 + 3:
    case 56 + 4:
    case 56 + 5:
    case 56 + 7:
    p = R8[op & 7];
    fc = *p & 1;
    FS(0, *p /= 2, 0, 0, fc);
    break;
    case 70:
    case 70 + 8:
    case 70 + 16:
    case 70 + 24:
    case 70 + 32:
    case 70 + 32 + 8:
    case 70 + 32 + 16:
    case 70 + 32 + 24:
    u = r(HL);
    goto BITU;
    case 64:
    case 64 + 8:
    case 64 + 16:
    case 64 + 24:
    case 64 + 32:
    case 64 + 40:
    case 64 + 56:
    case 64 + 1:
    case 64 + 1 + 8:
    case 64 + 1 + 16:
    case 64 + 1 + 24:
    case 64 + 1 + 32:
    case 64 + 1 + 40:
    case 64 + 1 + 56:
    case 64 + 2:
    case 64 + 2 + 8:
    case 64 + 2 + 16:
    case 64 + 2 + 24:
    case 64 + 2 + 32:
    case 64 + 2 + 40:
    case 64 + 2 + 56:
    case 64 + 3:
    case 64 + 3 + 8:
    case 64 + 3 + 16:
    case 64 + 3 + 24:
    case 64 + 3 + 32:
    case 64 + 3 + 40:
    case 64 + 3 + 56:
    case 64 + 4:
    case 64 + 4 + 8:
    case 64 + 4 + 16:
    case 64 + 4 + 24:
    case 64 + 4 + 32:
    case 64 + 4 + 40:
    case 64 + 4 + 56:
    case 64 + 5:
    case 64 + 5 + 8:
    case 64 + 5 + 16:
    case 64 + 5 + 24:
    case 64 + 5 + 32:
    case 64 + 5 + 40:
    case 64 + 5 + 56:
    case 64 + 7:
    case 64 + 7 + 8:
    case 64 + 7 + 16:
    case 64 + 7 + 24:
    case 64 + 7 + 32:
    case 64 + 7 + 40:
    case 64 + 7 + 56:
    case 64 + 48:
    case 64 + 48 + 1:
    case 64 + 48 + 2:
    case 64 + 48 + 3:
    case 64 + 48 + 4:
    case 64 + 48 + 5:
    case 64 + 48 + 7:
    p = R8[op & 7];
    u = *p;
    BITU:
    FS(16, u & (1 << ((op - 64) / 8)), 0, 1, 0);
    break;
    case 134:
    case 134 + 8:
    case 134 + 16:
    case 134 + 24:
    case 134 + 32:
    case 134 + 32 + 8:
    case 134 + 32 + 16:
    case 134 + 32 + 24:
    w(HL, r(HL) & ~(1 << ((op - 134) / 8)));
    break;
    case 128:
    case 128 + 8:
    case 128 + 16:
    case 128 + 24:
    case 128 + 32:
    case 128 + 40:
    case 128 + 56:
    case 128 + 1:
    case 128 + 1 + 8:
    case 128 + 1 + 16:
    case 128 + 1 + 24:
    case 128 + 1 + 32:
    case 128 + 1 + 40:
    case 128 + 1 + 56:
    case 128 + 2:
    case 128 + 2 + 8:
    case 128 + 2 + 16:
    case 128 + 2 + 24:
    case 128 + 2 + 32:
    case 128 + 2 + 40:
    case 128 + 2 + 56:
    case 128 + 3:
    case 128 + 3 + 8:
    case 128 + 3 + 16:
    case 128 + 3 + 24:
    case 128 + 3 + 32:
    case 128 + 3 + 40:
    case 128 + 3 + 56:
    case 128 + 4:
    case 128 + 4 + 8:
    case 128 + 4 + 16:
    case 128 + 4 + 24:
    case 128 + 4 + 32:
    case 128 + 4 + 40:
    case 128 + 4 + 56:
    case 128 + 5:
    case 128 + 5 + 8:
    case 128 + 5 + 16:
    case 128 + 5 + 24:
    case 128 + 5 + 32:
    case 128 + 5 + 40:
    case 128 + 5 + 56:
    case 128 + 7:
    case 128 + 7 + 8:
    case 128 + 7 + 16:
    case 128 + 7 + 24:
    case 128 + 7 + 32:
    case 128 + 7 + 40:
    case 128 + 7 + 56:
    case 128 + 48:
    case 128 + 48 + 1:
    case 128 + 48 + 2:
    case 128 + 48 + 3:
    case 128 + 48 + 4:
    case 128 + 48 + 5:
    case 128 + 48 + 7:
    p = R8[op & 7];
    *p &= ~(1 << ((op - 128) / 8));
    break;
    case 198:
    case 198 + 8:
    case 198 + 16:
    case 198 + 24:
    case 198 + 32:
    case 198 + 32 + 8:
    case 198 + 32 + 16:
    case 198 + 32 + 24:
    w(HL, r(HL) | (1 << ((op - 198) / 8)));
    break;
    case 192:
    case 192 + 8:
    case 192 + 16:
    case 192 + 24:
    case 192 + 32:
    case 192 + 40:
    case 192 + 56:
    case 192 + 1:
    case 192 + 1 + 8:
    case 192 + 1 + 16:
    case 192 + 1 + 24:
    case 192 + 1 + 32:
    case 192 + 1 + 40:
    case 192 + 1 + 56:
    case 192 + 2:
    case 192 + 2 + 8:
    case 192 + 2 + 16:
    case 192 + 2 + 24:
    case 192 + 2 + 32:
    case 192 + 2 + 40:
    case 192 + 2 + 56:
    case 192 + 3:
    case 192 + 3 + 8:
    case 192 + 3 + 16:
    case 192 + 3 + 24:
    case 192 + 3 + 32:
    case 192 + 3 + 40:
    case 192 + 3 + 56:
    case 192 + 4:
    case 192 + 4 + 8:
    case 192 + 4 + 16:
    case 192 + 4 + 24:
    case 192 + 4 + 32:
    case 192 + 4 + 40:
    case 192 + 4 + 56:
    case 192 + 5:
    case 192 + 5 + 8:
    case 192 + 5 + 16:
    case 192 + 5 + 24:
    case 192 + 5 + 32:
    case 192 + 5 + 40:
    case 192 + 5 + 56:
    case 192 + 7:
    case 192 + 7 + 8:
    case 192 + 7 + 16:
    case 192 + 7 + 24:
    case 192 + 7 + 32:
    case 192 + 7 + 40:
    case 192 + 7 + 56:
    case 192 + 48:
    case 192 + 48 + 1:
    case 192 + 48 + 2:
    case 192 + 48 + 3:
    case 192 + 48 + 4:
    case 192 + 48 + 5:
    case 192 + 48 + 7:
    p = R8[op & 7];
    *p |= 1 << ((op - 192) / 8);
    break;
    default:
    return op;
    switch (opcode = read8_pc()) {
    OP7_PTR(0)
    *ptr8 += *ptr8 + (carry = *ptr8 / 128 % 2);
    goto CARRY_ZERO_FLAGS_PTR;

    OP7_PTR(8)
    *ptr8 = (carry = *ptr8 & 1) * 128 + *ptr8 / 2;
    goto CARRY_ZERO_FLAGS_PTR;

    case 14:
    tmp8 = read8();
    carry = tmp8 & 1;
    write8(tmp8 = carry * 128 + tmp8 / 2);
    goto CARRY_ZERO_FLAGS_U;

    OP7_PTR(16)
    carry = *ptr8 >> 7;
    *ptr8 = *ptr8 * 2 + F / 16 % 2;
    goto CARRY_ZERO_FLAGS_PTR;

    OP7_PTR(24)
    carry = *ptr8 & 1;
    *ptr8 = *ptr8 / 2 + ((F * 8) & 128);
    goto CARRY_ZERO_FLAGS_PTR;

    OP7_PTR(32)
    carry = *ptr8 / 128 % 2;
    *ptr8 *= 2;
    goto CARRY_ZERO_FLAGS_PTR;

    OP7_PTR(40)
    carry = *ptr8 & 1;
    *ptr8 = (int8_t)*ptr8 >> 1;
    goto CARRY_ZERO_FLAGS_PTR;

    OP7_PTR(48)
    set_flags(0, *ptr8 = *ptr8 * 16 + *ptr8 / 16, 0, 0, 0);
    break;

    case 54:
    tmp8 = read8();
    carry = 0;
    write8(tmp8 = tmp8 * 16 + tmp8 / 16);
    CARRY_ZERO_FLAGS_U:
    set_flags(0, tmp8, 0, 0, carry);
    break;

    OP7_PTR(56)
    carry = *ptr8 & 1;
    *ptr8 /= 2;
    CARRY_ZERO_FLAGS_PTR:
    set_flags(0, *ptr8, 0, 0, carry);
    break;

    OP8_NX8_REL(70)
    tmp8 = read8();
    goto BIT_FLAGS;

    OP56_PTR_REL(64)
    tmp8 = *ptr8;
    BIT_FLAGS:
    set_flags(16, tmp8 & (1 << opcode_rel / 8), 0, 1, 0);
    break;

    OP8_NX8_REL(134)
    write8(read8() & ~(1 << opcode_rel / 8));
    break;

    OP56_PTR_REL(128)
    *ptr8 &= ~(1 << opcode_rel / 8);
    break;

    OP8_NX8_REL(198)
    write8(read8() | (1 << opcode_rel / 8));
    break;

    OP56_PTR_REL(192)
    *ptr8 |= 1 << opcode_rel / 8;
    break;
    }
    break;
    case 224:
    case 226:
    w(65280 + (op == 224 ? r8() : R[0]), A);

    case 224: case 226:
    write8(A, 65280 + (opcode == 224 ? read8_pc() : *reg8));
    break;

    case 233:
    PC = HL;
    break;

    case 234:
    w(r16(), A);
    write8(A, read16());
    break;
    case 240:
    A = r(65280 | r8());

    case 240: case 250:
    A = read8(opcode == 240 ? 65280 | read8_pc() : read16());
    break;
    case 243:
    case 251:
    IME = op != 243;

    case 243: case 251:
    IME = opcode != 243;
    break;

    case 248:
    u = r8();
    FS(0, 1, 0, (u8)SP + u > 255, SP % 16 + u % 16 > 15);
    HL = SP + (s8)u;
    T();
    tmp8 = read8_pc();
    set_flags(0, 1, 0, (uint8_t)SP + tmp8 > 255, SP % 16 + tmp8 % 16 > 15);
    HL = SP + (int8_t)tmp8;
    tick();
    break;

    case 249:
    SP = HL;
    T();
    break;
    case 250:
    A = r(r16());
    tick();
    break;
    default:
    return op;
    }
    for (DIV += cy - bcy; bcy++ < cy;)

    for (DIV += cycles - prev_cycles; prev_cycles++ != cycles;)
    if (LCDC & 128) {
    if (++dot == 1 && LY == 144)
    if (++ppu_dot == 1 && LY == 144)
    IF |= 1;
    if (dot == 456) {

    if (ppu_dot == 456) {
    if (LY < 144)
    for (int x = 160; --x >= 0;) {
    u16 win = LCDC & 32 && LY >= WY && x >= WX - 7,
    base = (LCDC & (win ? 64 : 8) ? 7 : 6) << 10;
    u8 mx = win ? x - WX + 7 : x + io[323],
    my = win ? LY - WY : LY + io[322];
    u16 tile = vram[base + my / 8 * 32 + mx / 8], p = 327;
    mx = (mx ^ 7) & 7;
    u8 *d = &vram[(LCDC & 16 ? tile : 256 + (s8)tile) * 16 +
    (my & 7) * 2];
    u8 c = (d[1] >> mx) % 2 * 2 + (d[0] >> mx) % 2;
    for (tmp = 160; --tmp >= 0;) {
    uint8_t is_window =
    LCDC & 32 && LY >= io[330] && tmp >= io[331] - 7,
    x_offset = is_window ? tmp - io[331] + 7 : tmp + io[323],
    y_offset = is_window ? LY - io[330] : LY + io[322];
    uint16_t tile = video_ram[((LCDC & (is_window ? 64 : 8) ? 7 : 6)
    << 10) +
    y_offset / 8 * 32 + x_offset / 8],
    palette_index = 0;

    x_offset = (x_offset ^ 7) & 7;

    uint8_t
    *tile_data =
    &video_ram[(LCDC & 16 ? tile : 256 + (int8_t)tile) * 16 +
    y_offset % 8 * 2],
    color = (tile_data[1] >> x_offset) % 2 * 2 +
    (*tile_data >> x_offset) % 2;

    if (LCDC & 2)
    for (u8 *o = io; o < io + 160; o += 4) {
    u8 dx = x - o[1] + 8, dy = LY - o[0] + 16;
    if (dx < 8 && dy < 8) {
    dx ^= o[3] & 32 ? 0 : 7;
    d = &vram[o[2] * 16 + (dy ^ (o[3] & 64 ? 7 : 0)) * 2];
    u8 nc = (d[1] >> dx) % 2 * 2 + (d[0] >> dx) % 2;
    if (!((o[3] & 128) && c) && nc) {
    c = nc;
    p += 1 + !!(o[3] & 8);
    for (uint8_t *sprite = io; sprite < io + 160; sprite += 4) {
    uint8_t sprite_x = tmp - sprite[1] + 8,
    sprite_y = LY - *sprite + 16;
    if (sprite_x < 8 && sprite_y < 8) {
    sprite_x ^= sprite[3] & 32 ? 0 : 7;

    tile_data =
    &video_ram[sprite[2] * 16 +
    (sprite_y ^ (sprite[3] & 64 ? 7 : 0)) * 2];
    uint8_t sprite_color = (tile_data[1] >> sprite_x) % 2 * 2 +
    (*tile_data >> sprite_x) % 2;

    if (!((sprite[3] & 128) && color) && sprite_color) {
    color = sprite_color;
    palette_index += 1 + !!(sprite[3] & 8);
    break;
    }
    }
    }
    fb[LY * 160 + x] = pal[(io[p] >> (2 * c)) & 3];

    frame_buffer[LY * 160 + tmp] =
    palette[(io[327 + palette_index] >> (2 * color)) % 4 +
    palette_index * 4];
    }

    if (LY == 144) {
    void *pixels;
    int pitch;
    SDL_LockTexture(St, 0, &pixels, &pitch);
    for (int y = 144; --y >= 0;)
    memcpy((u8 *)pixels + y * pitch, fb + y * 160, 640);
    SDL_LockTexture(St, 0, &pixels, &tmp);
    for (tmp2 = 144; --tmp2 >= 0;)
    memcpy((uint8_t *)pixels + tmp2 * tmp, frame_buffer + tmp2 * 160,
    640);
    SDL_UnlockTexture(St);
    SDL_RenderCopy(Sr, St, 0, 0);
    SDL_RenderPresent(Sr);
    SDL_Delay(0);
    SDL_Event e;
    while (SDL_PollEvent(&e))
    if (e.type == SDL_QUIT)
    return 0;
    }

    LY = (LY + 1) % 154;
    dot = 0;
    ppu_dot = 0;
    }
    } else {
    LY = dot = 0;
    }
    } else
    LY = ppu_dot = 0;
    }
    }
    }
  5. @binji binji created this gist May 28, 2021.
    870 changes: 870 additions & 0 deletions pokegb.cc
    View file
    Original file line number Diff line number Diff line change
    @@ -0,0 +1,870 @@
    using u8 = uint8_t;
    using s8 = int8_t;
    using u16 = uint16_t;
    using u32 = uint32_t;
    using u64 = uint64_t;

    u8 op, u, v, fc, *rom0, *rom1, io[512], vram[8192], wram[16384], *eram, *eram1,
    R[] = {19, 0, 216, 0, 77, 1, 176, 1, 254, 255}, &F = R[6], &A = R[7],
    *R8[] = {&R[1], &R[0], &R[3], &R[2], &R[5], &R[4], &F, &A}, &IF = io[271],
    &LCDC = io[320], &LY = io[324], &WY = io[330], &WX = io[331], &IE = io[511],
    IME = 0, halt = 0, FM[] = {128, 128, 16, 16}, FE[] = {0, 128, 0, 16}, *p;
    u8 const *Sk;
    u16 PC = 256, U, *RR = (u16 *)R, &HL = RR[2], &SP = RR[4],
    *RR2[] = {&RR[0], &RR[1], &HL, &SP}, *RR3[] = {&RR[0], &RR[1], &HL, &HL},
    &DIV = (u16 &)io[259], dot = 32, *P;
    u32 fb[23040], pal[] = {0xffffffff, 0xffaaaaaa, 0xff555555, 0xff000000};
    u64 bcy, cy = 0;
    int HA[] = {0, 0, 1, -1};

    void T() { cy += 4; }
    u8 mem(u16 a, u8 x, int w) {
    T();
    switch (a >> 12) {
    default:
    case 2 ... 3:
    if (w)
    rom1 = rom0 + ((x ? (x & 63) : 1) << 14);
    case 0 ... 1:
    return rom0[a];
    case 4 ... 5:
    if (w && x <= 3)
    eram1 = eram + (x << 13);
    case 6 ... 7:
    return rom1[a & 16383];
    case 8 ... 9:
    a &= 8191;
    if (w)
    vram[a] = x;
    return vram[a];
    case 10 ... 11:
    a &= 8191;
    if (w)
    eram1[a] = x;
    return eram1[a];
    case 15:
    if (a >= 65024) {
    if (w) {
    if (a == 65350)
    for (int i = 160; --i >= 0;)
    io[i] = mem(x * 256 + i, 0, 0);
    io[a & 511] = x;
    }
    if (a == 65280) {
    if (!(io[256] & 16))
    return ~(16 + Sk[81] * 8 + (Sk[82] * 4) + Sk[80] * 2 + Sk[79]);
    if (!(io[256] & 32))
    return ~(32 + Sk[40] * 8 + Sk[43] * 4 + Sk[29] * 2 + Sk[27]);
    return 255;
    }
    return io[a & 511];
    }
    case 12 ... 14:
    a &= 16383;
    if (w)
    wram[a] = x;
    return wram[a];
    }
    }
    u8 r(u16 a) { return mem(a, 0, 0); }
    void w(u16 a, u8 x) { mem(a, x, 1); }
    void FS(u8 M, int Z, int N, int H, int C) {
    F = (F & M) | (!Z * 128 + N * 64 + H * 32 + C * 16);
    }
    u8 r8() { return r(PC++); }
    u16 r16(u16 &R = PC) {
    u = r(R++);
    return r(R++) * 256 + u;
    }
    void push(u16 x) {
    w(--SP, x >> 8);
    w(--SP, x);
    T();
    }

    int main() {
    rom1 =
    (rom0 = (u8 *)mmap(0, 1048576, 0x1, 0x01, open("rom.gb", 00), 0)) + 32768;
    int sav = open("rom.sav", 0100 | 02, 0666);
    ftruncate(sav, 32768);
    eram1 = eram = (u8 *)mmap(0, 32768, 0x1 | 0x2, 0x01, sav, 0);
    LCDC = 145;
    DIV = 44032;
    SDL_Init((0x00000001u | 0x00000010u | 0x00000020u | 0x00004000u |
    0x00000200u | 0x00001000u | 0x00002000u | 0x00008000u));
    SDL_Window *Sw =
    SDL_CreateWindow("pokegb", 100, 100, 160 * 5, 144 * 5, SDL_WINDOW_SHOWN);
    SDL_Renderer *Sr = SDL_CreateRenderer(Sw, -1, 0);
    SDL_Texture *St = SDL_CreateTexture(Sr, SDL_PIXELFORMAT_RGBA32,
    SDL_TEXTUREACCESS_STREAMING, 160, 144);
    Sk = SDL_GetKeyboardState(0);

    while (1) {
    bcy = cy;
    if (IME & IF & IE) {
    IF = halt = IME = 0;
    cy += 8;
    fc = 1;
    U = 64;
    goto CALLU;
    } else if (halt)
    T();
    else
    switch ((op = r8())) {
    case 0:
    break;
    case 1:
    case 1 + 16:
    case 1 + 32:
    case 1 + 48:
    *RR2[op >> 4] = r16();
    break;
    case 2:
    case 2 + 16:
    case 2 + 32:
    case 2 + 48:
    w(*RR3[op >> 4], A);
    HL += HA[(op - 2) / 16];
    break;
    case 3:
    case 3 + 16:
    case 3 + 32:
    case 3 + 48:
    (*RR2[op >> 4])++;
    T();
    break;
    case 4:
    case 4 + 8:
    case 4 + 16:
    case 4 + 24:
    case 4 + 32:
    case 4 + 40:
    case 4 + 56:
    p = R8[(op - 4) / 8];
    ++(*p);
    FS(16, *p, 0, !(*p & 15), 0);
    break;
    case 5:
    case 5 + 8:
    case 5 + 16:
    case 5 + 24:
    case 5 + 32:
    case 5 + 40:
    case 5 + 56:
    p = R8[(op - 5) / 8];
    --(*p);
    FS(16, *p, 1, (*p & 15) == 15, 0);
    break;
    case 6:
    case 6 + 8:
    case 6 + 16:
    case 6 + 24:
    case 6 + 32:
    case 6 + 40:
    case 6 + 56:
    p = R8[(op - 6) / 8];
    *p = r8();
    break;
    case 7:
    A += A + (fc = A >> 7);
    FS(0, 1, 0, 0, fc);
    break;
    case 9:
    case 9 + 16:
    case 9 + 32:
    case 9 + 48:
    P = RR2[op >> 4];
    FS(128, 1, 0, HL % 4096 + *P % 4096 > 4095, HL + *P > 65535);
    HL += *P;
    T();
    break;
    case 10:
    case 10 + 16:
    case 10 + 32:
    case 10 + 48:
    A = r(*RR3[op >> 4]);
    HL += HA[(op - 10) / 16];
    break;
    case 11:
    case 11 + 16:
    case 11 + 32:
    case 11 + 48:
    (*RR2[op >> 4])--;
    T();
    break;
    case 15:
    A = (fc = A & 1) * 128 + A / 2;
    FS(0, 1, 0, 0, fc);
    break;
    case 23:
    fc = A >> 7;
    A += A + (F >> 4) % 2;
    FS(0, 1, 0, 0, fc);
    break;
    case 32:
    case 32 + 8:
    case 32 + 16:
    case 32 + 24:
    case 24:
    fc = (op == 24) || (F & FM[(op - 32) / 8]) == FE[(op - 32) / 8];
    u = r8();
    if (fc)
    PC += (s8)u, T();
    break;
    case 39:
    fc = u = 0;
    if (F & 32 || (!(F & 64) && A % 15 > 9))
    u = 6;
    if (F & 16 || (!(F & 64) && A > 153))
    u |= 96, fc = 1;
    FS(65, A += (F & 64) ? -u : u, 0, 0, fc);
    break;
    case 47:
    A = ~A;
    FS(129, 1, 1, 1, 0);
    break;
    case 52:
    u = r(HL) + 1;
    FS(16, u, 0, !(u & 15), 0);
    w(HL, u);
    break;
    case 53:
    u = r(HL) - 1;
    FS(16, u, 1, u % 16 == 15, 0);
    w(HL, u);
    break;
    case 54:
    w(HL, r8());
    break;
    case 55:
    FS(128, 1, 0, 0, 1);
    break;
    case 63:
    FS(128, 1, 0, 0, !(F & 16));
    break;
    case 70:
    case 70 + 8:
    case 70 + 16:
    case 70 + 24:
    case 70 + 32:
    case 70 + 40:
    case 70 + 56:
    p = R8[(op - 70) / 8];
    *p = r(HL);
    break;
    case 64:
    case 64 + 8:
    case 64 + 16:
    case 64 + 24:
    case 64 + 32:
    case 64 + 40:
    case 64 + 56:
    case 64 + 1:
    case 64 + 1 + 8:
    case 64 + 1 + 16:
    case 64 + 1 + 24:
    case 64 + 1 + 32:
    case 64 + 1 + 40:
    case 64 + 1 + 56:
    case 64 + 2:
    case 64 + 2 + 8:
    case 64 + 2 + 16:
    case 64 + 2 + 24:
    case 64 + 2 + 32:
    case 64 + 2 + 40:
    case 64 + 2 + 56:
    case 64 + 3:
    case 64 + 3 + 8:
    case 64 + 3 + 16:
    case 64 + 3 + 24:
    case 64 + 3 + 32:
    case 64 + 3 + 40:
    case 64 + 3 + 56:
    case 64 + 4:
    case 64 + 4 + 8:
    case 64 + 4 + 16:
    case 64 + 4 + 24:
    case 64 + 4 + 32:
    case 64 + 4 + 40:
    case 64 + 4 + 56:
    case 64 + 5:
    case 64 + 5 + 8:
    case 64 + 5 + 16:
    case 64 + 5 + 24:
    case 64 + 5 + 32:
    case 64 + 5 + 40:
    case 64 + 5 + 56:
    case 64 + 7:
    case 64 + 7 + 8:
    case 64 + 7 + 16:
    case 64 + 7 + 24:
    case 64 + 7 + 32:
    case 64 + 7 + 40:
    case 64 + 7 + 56:
    *R8[(op - 64) / 8] = *R8[op & 7];
    break;
    case 112:
    case 112 + 1:
    case 112 + 2:
    case 112 + 3:
    case 112 + 4:
    case 112 + 5:
    case 112 + 7:
    p = R8[op & 7];
    w(HL, *p);
    break;
    case 118:
    halt = 1;
    break;
    case 128 + 6:
    case 128 + 70:
    case 128:
    case 128 + 1:
    case 128 + 2:
    case 128 + 3:
    case 128 + 4:
    case 128 + 5:
    case 128 + 7:
    p = R8[op & 7];
    v = (op == 128 + 6) ? r(HL) : (op == 128 + 70) ? r8() : *p;
    fc = 0;
    goto ADD;
    case 136 + 6:
    case 136 + 70:
    case 136:
    case 136 + 1:
    case 136 + 2:
    case 136 + 3:
    case 136 + 4:
    case 136 + 5:
    case 136 + 7:
    p = R8[op & 7];
    v = (op == 136 + 6) ? r(HL) : (op == 136 + 70) ? r8() : *p;
    fc = (F >> 4) & 1;
    ADD:
    FS(0, u = A + v + fc, 0, A % 16 + v % 16 + fc > 15, A + v + fc > 255);
    A = u;
    break;
    case 144 + 6:
    case 144 + 70:
    case 144:
    case 144 + 1:
    case 144 + 2:
    case 144 + 3:
    case 144 + 4:
    case 144 + 5:
    case 144 + 7:
    p = R8[op & 7];
    v = (op == 144 + 6) ? r(HL) : (op == 144 + 70) ? r8() : *p;
    fc = 0;
    goto SUB;
    case 152 + 6:
    case 152 + 70:
    case 152:
    case 152 + 1:
    case 152 + 2:
    case 152 + 3:
    case 152 + 4:
    case 152 + 5:
    case 152 + 7:
    p = R8[op & 7];
    v = (op == 152 + 6) ? r(HL) : (op == 152 + 70) ? r8() : *p;
    fc = (F >> 4) & 1;
    SUB:
    FS(0, u = A - v - fc, 1, A % 16 < v % 16 + fc, A < v + fc);
    A = u;
    break;
    case 160 + 6:
    case 160 + 70:
    case 160:
    case 160 + 1:
    case 160 + 2:
    case 160 + 3:
    case 160 + 4:
    case 160 + 5:
    case 160 + 7:
    p = R8[op & 7];
    v = (op == 160 + 6) ? r(HL) : (op == 160 + 70) ? r8() : *p;
    FS(0, A &= v, 0, 1, 0);
    break;
    case 168 + 6:
    case 168 + 70:
    case 168:
    case 168 + 1:
    case 168 + 2:
    case 168 + 3:
    case 168 + 4:
    case 168 + 5:
    case 168 + 7:
    p = R8[op & 7];
    v = (op == 168 + 6) ? r(HL) : (op == 168 + 70) ? r8() : *p;
    FS(0, A ^= v, 0, 0, 0);
    break;
    case 176 + 6:
    case 176 + 70:
    case 176:
    case 176 + 1:
    case 176 + 2:
    case 176 + 3:
    case 176 + 4:
    case 176 + 5:
    case 176 + 7:
    p = R8[op & 7];
    v = (op == 176 + 6) ? r(HL) : (op == 176 + 70) ? r8() : *p;
    FS(0, A |= v, 0, 0, 0);
    break;
    case 184 + 6:
    case 184 + 70:
    case 184:
    case 184 + 1:
    case 184 + 2:
    case 184 + 3:
    case 184 + 4:
    case 184 + 5:
    case 184 + 7:
    p = R8[op & 7];
    v = (op == 184 + 6) ? r(HL) : (op == 184 + 70) ? r8() : *p;
    FS(0, A != v, 1, A % 16 < v % 16, A < v);
    break;
    case 217:
    fc = IME = 1;
    goto RET;
    case 192:
    case 192 + 8:
    case 192 + 16:
    case 192 + 24:
    case 201:
    fc = (op == 201) || (F & FM[(op - 192) / 8]) == FE[(op - 192) / 8];
    RET:
    T();
    if (fc)
    PC = r16(SP);
    break;
    case 193:
    case 193 + 16:
    case 193 + 32:
    case 193 + 48:
    RR[(op - 193) >> 4] = r16(SP);
    break;
    case 194:
    case 194 + 8:
    case 194 + 16:
    case 194 + 24:
    case 195:
    fc = (op == 195) || (F & FM[(op - 194) / 8]) == FE[(op - 194) / 8];
    U = r16();
    if (fc)
    PC = U, T();
    break;
    case 196:
    case 196 + 8:
    case 196 + 16:
    case 196 + 24:
    case 205:
    fc = (op == 205) || (F & FM[(op - 196) / 8]) == FE[(op - 196) / 8];
    U = r16();
    CALLU:
    if (fc)
    push(PC), PC = U;
    break;
    case 197:
    case 197 + 16:
    case 197 + 32:
    case 197 + 48:
    push(RR[(op - 197) >> 4]);
    break;
    case 203:
    switch ((op = r8())) {
    case 7:
    A += A + (fc = (A >> 7) & 1);
    FS(0, A, 0, 0, fc);
    break;
    case 8:
    case 8 + 1:
    case 8 + 2:
    case 8 + 3:
    case 8 + 4:
    case 8 + 5:
    case 8 + 7:
    p = R8[op & 7];
    *p = (fc = *p & 1) * 128 + *p / 2;
    FS(0, *p, 0, 0, fc);
    break;
    case 14:
    u = r(HL);
    fc = u & 1;
    u = fc * 128 + u / 2;
    FS(0, u, 0, 0, fc);
    w(HL, u);
    break;
    case 16:
    case 16 + 1:
    case 16 + 2:
    case 16 + 3:
    case 16 + 4:
    case 16 + 5:
    case 16 + 7:
    p = R8[op & 7];
    fc = *p >> 7;
    FS(0, *p = *p * 2 + (F >> 4) % 2, 0, 0, fc);
    break;
    case 24:
    case 24 + 1:
    case 24 + 2:
    case 24 + 3:
    case 24 + 4:
    case 24 + 5:
    case 24 + 7:
    p = R8[op & 7];
    fc = *p & 1;
    FS(0, *p = *p / 2 + ((F * 8) & 128), 0, 0, fc);
    break;
    case 32:
    case 32 + 1:
    case 32 + 2:
    case 32 + 3:
    case 32 + 4:
    case 32 + 5:
    case 32 + 7:
    p = R8[op & 7];
    fc = (*p >> 7) & 1;
    FS(0, *p *= 2, 0, 0, fc);
    break;
    case 40:
    case 40 + 1:
    case 40 + 2:
    case 40 + 3:
    case 40 + 4:
    case 40 + 5:
    case 40 + 7:
    p = R8[op & 7];
    fc = *p & 1;
    FS(0, *p = (s8)*p >> 1, 0, 0, fc);
    break;
    case 48:
    case 48 + 1:
    case 48 + 2:
    case 48 + 3:
    case 48 + 4:
    case 48 + 5:
    case 48 + 7:
    p = R8[op & 7];
    FS(0, *p = *p * 16 + *p / 16, 0, 0, 0);
    break;
    case 54:
    u = r(HL);
    FS(0, u = u * 16 + u / 16, 0, 0, 0);
    w(HL, u);
    break;
    case 56:
    case 56 + 1:
    case 56 + 2:
    case 56 + 3:
    case 56 + 4:
    case 56 + 5:
    case 56 + 7:
    p = R8[op & 7];
    fc = *p & 1;
    FS(0, *p /= 2, 0, 0, fc);
    break;
    case 70:
    case 70 + 8:
    case 70 + 16:
    case 70 + 24:
    case 70 + 32:
    case 70 + 32 + 8:
    case 70 + 32 + 16:
    case 70 + 32 + 24:
    u = r(HL);
    goto BITU;
    case 64:
    case 64 + 8:
    case 64 + 16:
    case 64 + 24:
    case 64 + 32:
    case 64 + 40:
    case 64 + 56:
    case 64 + 1:
    case 64 + 1 + 8:
    case 64 + 1 + 16:
    case 64 + 1 + 24:
    case 64 + 1 + 32:
    case 64 + 1 + 40:
    case 64 + 1 + 56:
    case 64 + 2:
    case 64 + 2 + 8:
    case 64 + 2 + 16:
    case 64 + 2 + 24:
    case 64 + 2 + 32:
    case 64 + 2 + 40:
    case 64 + 2 + 56:
    case 64 + 3:
    case 64 + 3 + 8:
    case 64 + 3 + 16:
    case 64 + 3 + 24:
    case 64 + 3 + 32:
    case 64 + 3 + 40:
    case 64 + 3 + 56:
    case 64 + 4:
    case 64 + 4 + 8:
    case 64 + 4 + 16:
    case 64 + 4 + 24:
    case 64 + 4 + 32:
    case 64 + 4 + 40:
    case 64 + 4 + 56:
    case 64 + 5:
    case 64 + 5 + 8:
    case 64 + 5 + 16:
    case 64 + 5 + 24:
    case 64 + 5 + 32:
    case 64 + 5 + 40:
    case 64 + 5 + 56:
    case 64 + 7:
    case 64 + 7 + 8:
    case 64 + 7 + 16:
    case 64 + 7 + 24:
    case 64 + 7 + 32:
    case 64 + 7 + 40:
    case 64 + 7 + 56:
    case 64 + 48:
    case 64 + 48 + 1:
    case 64 + 48 + 2:
    case 64 + 48 + 3:
    case 64 + 48 + 4:
    case 64 + 48 + 5:
    case 64 + 48 + 7:
    p = R8[op & 7];
    u = *p;
    BITU:
    FS(16, u & (1 << ((op - 64) / 8)), 0, 1, 0);
    break;
    case 134:
    case 134 + 8:
    case 134 + 16:
    case 134 + 24:
    case 134 + 32:
    case 134 + 32 + 8:
    case 134 + 32 + 16:
    case 134 + 32 + 24:
    w(HL, r(HL) & ~(1 << ((op - 134) / 8)));
    break;
    case 128:
    case 128 + 8:
    case 128 + 16:
    case 128 + 24:
    case 128 + 32:
    case 128 + 40:
    case 128 + 56:
    case 128 + 1:
    case 128 + 1 + 8:
    case 128 + 1 + 16:
    case 128 + 1 + 24:
    case 128 + 1 + 32:
    case 128 + 1 + 40:
    case 128 + 1 + 56:
    case 128 + 2:
    case 128 + 2 + 8:
    case 128 + 2 + 16:
    case 128 + 2 + 24:
    case 128 + 2 + 32:
    case 128 + 2 + 40:
    case 128 + 2 + 56:
    case 128 + 3:
    case 128 + 3 + 8:
    case 128 + 3 + 16:
    case 128 + 3 + 24:
    case 128 + 3 + 32:
    case 128 + 3 + 40:
    case 128 + 3 + 56:
    case 128 + 4:
    case 128 + 4 + 8:
    case 128 + 4 + 16:
    case 128 + 4 + 24:
    case 128 + 4 + 32:
    case 128 + 4 + 40:
    case 128 + 4 + 56:
    case 128 + 5:
    case 128 + 5 + 8:
    case 128 + 5 + 16:
    case 128 + 5 + 24:
    case 128 + 5 + 32:
    case 128 + 5 + 40:
    case 128 + 5 + 56:
    case 128 + 7:
    case 128 + 7 + 8:
    case 128 + 7 + 16:
    case 128 + 7 + 24:
    case 128 + 7 + 32:
    case 128 + 7 + 40:
    case 128 + 7 + 56:
    case 128 + 48:
    case 128 + 48 + 1:
    case 128 + 48 + 2:
    case 128 + 48 + 3:
    case 128 + 48 + 4:
    case 128 + 48 + 5:
    case 128 + 48 + 7:
    p = R8[op & 7];
    *p &= ~(1 << ((op - 128) / 8));
    break;
    case 198:
    case 198 + 8:
    case 198 + 16:
    case 198 + 24:
    case 198 + 32:
    case 198 + 32 + 8:
    case 198 + 32 + 16:
    case 198 + 32 + 24:
    w(HL, r(HL) | (1 << ((op - 198) / 8)));
    break;
    case 192:
    case 192 + 8:
    case 192 + 16:
    case 192 + 24:
    case 192 + 32:
    case 192 + 40:
    case 192 + 56:
    case 192 + 1:
    case 192 + 1 + 8:
    case 192 + 1 + 16:
    case 192 + 1 + 24:
    case 192 + 1 + 32:
    case 192 + 1 + 40:
    case 192 + 1 + 56:
    case 192 + 2:
    case 192 + 2 + 8:
    case 192 + 2 + 16:
    case 192 + 2 + 24:
    case 192 + 2 + 32:
    case 192 + 2 + 40:
    case 192 + 2 + 56:
    case 192 + 3:
    case 192 + 3 + 8:
    case 192 + 3 + 16:
    case 192 + 3 + 24:
    case 192 + 3 + 32:
    case 192 + 3 + 40:
    case 192 + 3 + 56:
    case 192 + 4:
    case 192 + 4 + 8:
    case 192 + 4 + 16:
    case 192 + 4 + 24:
    case 192 + 4 + 32:
    case 192 + 4 + 40:
    case 192 + 4 + 56:
    case 192 + 5:
    case 192 + 5 + 8:
    case 192 + 5 + 16:
    case 192 + 5 + 24:
    case 192 + 5 + 32:
    case 192 + 5 + 40:
    case 192 + 5 + 56:
    case 192 + 7:
    case 192 + 7 + 8:
    case 192 + 7 + 16:
    case 192 + 7 + 24:
    case 192 + 7 + 32:
    case 192 + 7 + 40:
    case 192 + 7 + 56:
    case 192 + 48:
    case 192 + 48 + 1:
    case 192 + 48 + 2:
    case 192 + 48 + 3:
    case 192 + 48 + 4:
    case 192 + 48 + 5:
    case 192 + 48 + 7:
    p = R8[op & 7];
    *p |= 1 << ((op - 192) / 8);
    break;
    default:
    return op;
    }
    break;
    case 224:
    case 226:
    w(65280 + (op == 224 ? r8() : R[0]), A);
    break;
    case 233:
    PC = HL;
    break;
    case 234:
    w(r16(), A);
    break;
    case 240:
    A = r(65280 | r8());
    break;
    case 243:
    case 251:
    IME = op != 243;
    break;
    case 248:
    u = r8();
    FS(0, 1, 0, (u8)SP + u > 255, SP % 16 + u % 16 > 15);
    HL = SP + (s8)u;
    T();
    break;
    case 249:
    SP = HL;
    T();
    break;
    case 250:
    A = r(r16());
    break;
    default:
    return op;
    }
    for (DIV += cy - bcy; bcy++ < cy;)
    if (LCDC & 128) {
    if (++dot == 1 && LY == 144)
    IF |= 1;
    if (dot == 456) {
    if (LY < 144)
    for (int x = 160; --x >= 0;) {
    u16 win = LCDC & 32 && LY >= WY && x >= WX - 7,
    base = (LCDC & (win ? 64 : 8) ? 7 : 6) << 10;
    u8 mx = win ? x - WX + 7 : x + io[323],
    my = win ? LY - WY : LY + io[322];
    u16 tile = vram[base + my / 8 * 32 + mx / 8], p = 327;
    mx = (mx ^ 7) & 7;
    u8 *d = &vram[(LCDC & 16 ? tile : 256 + (s8)tile) * 16 +
    (my & 7) * 2];
    u8 c = (d[1] >> mx) % 2 * 2 + (d[0] >> mx) % 2;
    if (LCDC & 2)
    for (u8 *o = io; o < io + 160; o += 4) {
    u8 dx = x - o[1] + 8, dy = LY - o[0] + 16;
    if (dx < 8 && dy < 8) {
    dx ^= o[3] & 32 ? 0 : 7;
    d = &vram[o[2] * 16 + (dy ^ (o[3] & 64 ? 7 : 0)) * 2];
    u8 nc = (d[1] >> dx) % 2 * 2 + (d[0] >> dx) % 2;
    if (!((o[3] & 128) && c) && nc) {
    c = nc;
    p += 1 + !!(o[3] & 8);
    break;
    }
    }
    }
    fb[LY * 160 + x] = pal[(io[p] >> (2 * c)) & 3];
    }
    if (LY == 144) {
    void *pixels;
    int pitch;
    SDL_LockTexture(St, 0, &pixels, &pitch);
    for (int y = 144; --y >= 0;)
    memcpy((u8 *)pixels + y * pitch, fb + y * 160, 640);
    SDL_UnlockTexture(St);
    SDL_RenderCopy(Sr, St, 0, 0);
    SDL_RenderPresent(Sr);
    SDL_Delay(0);
    SDL_Event e;
    while (SDL_PollEvent(&e))
    if (e.type == SDL_QUIT)
    return 0;
    }
    LY = (LY + 1) % 154;
    dot = 0;
    }
    } else {
    LY = dot = 0;
    }
    }
    }