add gb_config_t, gb_disasm(), refactor logic, misc cleanups

1 files changed, 358 insertions, 392 deletions

diff --git a/ops.yaml b/ops.yaml
index 9266820..b1d2ad3 100644
--- a/ops.yaml
+++ b/ops.yaml
@@ -761,6 +761,7 @@ ops:
         c:
       code: |
         cpu_wb(ctx, RB_A, cpu_rb(ctx, RB_A) ^ 0xFF);
+        cpu_wf(ctx, F_N | F_H, F_N | F_H);
     - id: JR NC, r8
       hex: 0x30
       cat: "jumps/calls"
@@ -3746,7 +3747,8 @@ ops:
         h:
         c:
       code: |
-        cpu_ww(ctx, RW_PC, mmu_rw(ctx, cpu_rw(ctx, RW_HL)));
+        // cpu_ww(ctx, RW_PC, mmu_rw(ctx, cpu_rw(ctx, RW_HL)));
+        cpu_ww(ctx, RW_PC, cpu_rw(ctx, RW_HL));
     - id: LD (a16), A
       hex: 0xEA
       cat: "8-bit load/store/move"
@@ -8073,6 +8075,7 @@ templates:
     #include <string.h>   // memset()
     #include <stdint.h>   // uint8_t ,etc
     #include <stdbool.h>  // bool
+    #include <stdio.h>    // snprintf()
     #include "gb.h"
 
     #define UNUSED(a) ((void) (a))
@@ -8227,7 +8230,7 @@ templates:
         (((ctx->mmu.mbc1.rom_ram_mode ? 0 : ctx->mmu.mbc1.ram_bank) & 0x3) << 19) |
 
         // B: rom_bank (5 bits, low bit set)
-        (((ctx->mmu.mbc1.rom_bank & 0x1F) | 0x1) << 15) |
+        (((ctx->mmu.mbc1.rom_bank & 0x1F) | 0x1) << 14) |
 
         // C: address (14 bits)
         ((addr - 0x4000) & 0x3FFF)
@@ -8262,9 +8265,11 @@ templates:
       case 0x6000:
       case 0x7000:
         // ROM1 (16k, banked)
-        // FIXME: clamp to rom size?
-        // mbc1 rom1 banking (TCAGBD.pdf, 11.2.2)
-        return ctx->mmu.rom[mbc1_rom_get_addr(ctx, addr)];
+        {
+          // mbc1 rom1 banking (TCAGBD.pdf, 11.2.2)
+          const uint32_t addr_ofs = mbc1_rom_get_addr(ctx, addr);
+          return addr_ofs < ctx->mmu.rom_size ? ctx->mmu.rom[addr_ofs] : 0;
+        }
       default:
         // never reached
         return 0;
@@ -8285,13 +8290,24 @@ templates:
         break;
       case 0x2000:
       case 0x3000:
-        // set rom bank
         ctx->mmu.mbc1.rom_bank = val & 0x1F;
+
+        if (ctx->config && ctx->config->on_set_rom_bank) {
+          // notify callback
+          ctx->config->on_set_rom_bank(ctx, val & 0x1F);
+        }
+
         break;
       case 0x4000:
       case 0x5000:
         // set ram_bank/upper_rom_bank bits
-        ctx->mmu.mbc1.ram_bank = val & 0x3;
+        ctx->mmu.mbc1.ram_bank = val & 0x03;
+
+        if (ctx->config && ctx->config->on_set_ram_bank) {
+          // notify callback
+          ctx->config->on_set_ram_bank(ctx, val & 0x03);
+        }
+
         break;
       case 0x6000:
       case 0x7000:
@@ -8393,8 +8409,10 @@ templates:
       case 0x6000:
       case 0x7000:
         // ROM1 (16k, banked)
-        // FIXME: clamp to rom size?
-        return ctx->mmu.rom[mbc2_rom_get_addr(ctx, addr)];
+        {
+          const uint32_t addr_ofs = mbc2_rom_get_addr(ctx, addr);
+          return addr_ofs < ctx->mmu.rom_size ? ctx->mmu.rom[addr_ofs] : 0;
+        }
       default:
         // never reached
         return 0;
@@ -8633,6 +8651,9 @@ templates:
       }
     }
 
+    // forward reference for mmu_wp()
+    static void gpu_set_line(gb_t * const ctx, const uint8_t line);
+
     static void
     mmu_wp(
       gb_t * const ctx,
@@ -8700,8 +8721,9 @@ templates:
 
         break;
       case PORT_LY:
-        // FIXME: according to URL below writing to this register zeroes it:
+        // NOTE: according to URL below writing to this register zeroes it:
         // (src: http://gbdev.gg8.se/wiki/articles/Video_Display)
+        gpu_set_line(ctx, 0);
 
         break;
       case PORT_LYC:
@@ -8782,7 +8804,7 @@ templates:
       const gb_t * const ctx,
       const uint16_t addr
     ) {
-      switch (ctx->gpu.mode) {
+      switch (ctx->gpu.mode & 0x3) {
       case GPU_MODE_VRAM:
         return 0xFF;
       default:
@@ -8796,7 +8818,7 @@ templates:
       const gb_t * const ctx,
       const uint16_t addr
     ) {
-      switch (ctx->gpu.mode) {
+      switch (ctx->gpu.mode & 0x3) {
       case GPU_MODE_HBLANK:
       case GPU_MODE_VBLANK:
         // oam memory (160 bytes):
@@ -8932,7 +8954,7 @@ templates:
       const uint16_t addr,
       const uint8_t val
     ) {
-      switch (ctx->gpu.mode) {
+      switch (ctx->gpu.mode & 0x3) {
       case GPU_MODE_HBLANK:
       case GPU_MODE_VBLANK:
       case GPU_MODE_OAM:
@@ -8951,7 +8973,7 @@ templates:
       const uint16_t addr,
       const uint8_t val
     ) {
-      switch (ctx->gpu.mode) {
+      switch (ctx->gpu.mode & 0x3) {
       case GPU_MODE_HBLANK:
       case GPU_MODE_VBLANK:
         // oam memory (160 bytes):
@@ -9171,7 +9193,12 @@ templates:
       const gb_cpu_state_t state
     ) {
       ctx->cpu.state = state;
+
       UNUSED(pc);
+
+      if (ctx->config && ctx->config->on_set_cpu_state) {
+        ctx->config->on_set_cpu_state(ctx, state);
+      }
     }
 
     static void
@@ -9274,13 +9301,12 @@ templates:
     }
 
     static void
-    add_rb(
+    add(
       gb_t * const ctx,
-      const rb_t reg
+      const uint8_t n
     ) {
       // get old and new value
-      const uint8_t o = cpu_rb(ctx, RB_A),
-                    n = cpu_rb(ctx, reg);
+      const uint8_t o = cpu_rb(ctx, RB_A);
       const uint16_t v = o + n;
 
       // write value
@@ -9295,44 +9321,26 @@ templates:
     }
 
     static void
+    add_rb(
+      gb_t * const ctx,
+      const rb_t reg
+    ) {
+      add(ctx, cpu_rb(ctx, reg));
+    }
+
+    static void
     add_d8(
       gb_t * const ctx,
       const uint16_t addr
     ) {
-      // get old and new value
-      const uint8_t o = cpu_rb(ctx, RB_A),
-                    n = mmu_rb(ctx, addr);
-      const uint16_t v = o + n;
-
-      // write value
-      cpu_wb(ctx, RB_A, v);
-
-      // set flags
-      cpu_wf(ctx, F_Z | F_N | F_H | F_C, (
-        ((v & 0xFF) ? 0 : F_Z) |
-        ((((o & 0x0F) + (n & 0x0F)) & 0xF0) ? F_H : 0) |
-        ((v & 0x100) ? F_C : 0)
-      ));
+      add(ctx, mmu_rb(ctx, addr));
     }
 
     static void
     add_hl_ptr(
       gb_t * const ctx
     ) {
-      // get old and new value
-      const uint8_t o = cpu_rb(ctx, RB_A),
-                    n = mmu_rw(ctx, cpu_rw(ctx, RW_HL));
-      const uint16_t v = o + n;
-
-      // write value
-      cpu_wb(ctx, RB_A, v);
-
-      // set flags
-      cpu_wf(ctx, F_Z | F_N | F_H | F_C, (
-        ((v & 0xFF) ? 0 : F_Z) |
-        ((((o & 0x0F) + (n & 0x0F)) & 0xF0) ? F_H : 0) |
-        ((v & 0x100) ? F_C : 0)
-      ));
+      add(ctx, mmu_rb(ctx, cpu_rw(ctx, RW_HL)));
     }
 
     static void
@@ -9359,11 +9367,11 @@ templates:
     static void
     ld_hl_sp_r8(
       gb_t * const ctx,
-      const uint16_t addr
+      const uint16_t ofs_addr
     ) {
       // get src values
       const uint16_t a = cpu_rw(ctx, RW_SP);
-      const int8_t b = (int8_t) mmu_rb(ctx, addr);
+      const int8_t b = (int8_t) mmu_rb(ctx, ofs_addr);
       const uint16_t v = a + b;
 
       // write dst value
@@ -9377,13 +9385,12 @@ templates:
     }
 
     static void
-    adc_rb(
+    adc(
       gb_t * const ctx,
-      const rb_t reg
+      const uint8_t n
     ) {
-      // get old and new value
+      // get old value and carry flag
       const uint8_t o = cpu_rb(ctx, RB_A),
-                    n = cpu_rb(ctx, reg),
                     c = FLAG(ctx, C) ? 1 : 0;
       const uint16_t v = o + n + c;
 
@@ -9399,24 +9406,18 @@ templates:
     }
 
     static void
+    adc_rb(
+      gb_t * const ctx,
+      const rb_t reg
+    ) {
+      adc(ctx, cpu_rb(ctx, reg));
+    }
+
+    static void
     adc_hl_ptr(
       gb_t * const ctx
     ) {
-      // get old and new value
-      const uint8_t o = cpu_rb(ctx, RB_A),
-                    n = cpu_rb(ctx, mmu_rb(ctx, cpu_rw(ctx, RW_HL))),
-                    c = FLAG(ctx, C) ? 1 : 0;
-      const uint16_t v = o + n + c;
-
-      // write value
-      cpu_wb(ctx, RB_A, v);
-
-      // set flags
-      cpu_wf(ctx, F_Z | F_N | F_H | F_C, (
-        ((v & 0xFF) ? 0 : F_Z) |
-        ((((o & 0x0F) + (n & 0x0F) + c) & 0xF0) ? F_H : 0) |
-        ((v & 0x100) ? F_C : 0)
-      ));
+      adc(ctx, mmu_rb(ctx, cpu_rw(ctx, RW_HL)));
     }
 
     static void
@@ -9424,31 +9425,16 @@ templates:
       gb_t * const ctx,
       const uint16_t addr
     ) {
-      // get old and new value
-      const uint8_t o = cpu_rb(ctx, RB_A),
-                    n = mmu_rb(ctx, addr),
-                    c = FLAG(ctx, C) ? 1 : 0;
-      const uint16_t v = o + n + c;
-
-      // write value
-      cpu_wb(ctx, RB_A, v);
-
-      // set flags
-      cpu_wf(ctx, F_Z | F_N | F_H | F_C, (
-        ((v & 0xFF) ? 0 : F_Z) |
-        ((((o & 0x0F) + (n & 0x0F) + c) & 0xF0) ? F_H : 0) |
-        ((v & 0x100) ? F_C : 0)
-      ));
+      adc(ctx, mmu_rb(ctx, addr));
     }
 
     static void
-    sub_rb(
+    sub(
       gb_t * const ctx,
-      const rb_t reg
+      const uint8_t n
     ) {
       // get old and new value
       const uint8_t o = cpu_rb(ctx, RB_A),
-                    n = cpu_rb(ctx, reg),
                     v = o - n;
 
       // write value
@@ -9464,24 +9450,18 @@ templates:
     }
 
     static void
+    sub_rb(
+      gb_t * const ctx,
+      const rb_t reg
+    ) {
+      sub(ctx, cpu_rb(ctx, reg));
+    }
+
+    static void
     sub_hl_ptr(
       gb_t * const ctx
     ) {
-      // get old and new value
-      const uint8_t o = cpu_rb(ctx, RB_A),
-                    n = mmu_rb(ctx, cpu_rw(ctx, RW_HL)),
-                    v = o - n;
-
-      // write value
-      cpu_wb(ctx, RB_A, v);
-
-      // set flags
-      cpu_wf(ctx, F_Z | F_N | F_H | F_C, (
-        (v ? 0 : F_Z) |
-        F_N |
-        ((((o & 0x0F) - (n & 0x0F)) & 0xF0) ? F_H : 0) |
-        ((n > o) ? F_C : 0)
-      ));
+      sub(ctx, mmu_rb(ctx, cpu_rw(ctx, RW_HL)));
     }
 
     static void
@@ -9489,31 +9469,16 @@ templates:
       gb_t * const ctx,
       const uint16_t addr
     ) {
-      // get old and new value
-      const uint8_t o = cpu_rb(ctx, RB_A),
-                    n = mmu_rb(ctx, addr),
-                    v = o - n;
-
-      // write value
-      cpu_wb(ctx, RB_A, v);
-
-      // set flags
-      cpu_wf(ctx, F_Z | F_N | F_H | F_C, (
-        (v ? 0 : F_Z) |
-        F_N |
-        ((((o & 0x0F) - (n & 0x0F)) & 0xF0) ? F_H : 0) |
-        ((n > o) ? F_C : 0)
-      ));
+      sub(ctx, mmu_rb(ctx, addr));
     }
 
     static void
-    sbc_rb(
+    sbc(
       gb_t * const ctx,
-      const rb_t reg
+      const uint8_t n
     ) {
       // get old and new value
       const uint8_t o = cpu_rb(ctx, RB_A),
-                    n = cpu_rb(ctx, reg),
                     c = FLAG(ctx, C) ? 1 : 0,
                     v = o - (n + c);
 
@@ -9530,34 +9495,27 @@ templates:
     }
 
     static void
+    sbc_rb(
+      gb_t * const ctx,
+      const rb_t reg
+    ) {
+      sbc(ctx, cpu_rb(ctx, reg));
+    }
+
+    static void
     sbc_hl_ptr(
       gb_t * const ctx
     ) {
-      // get old and new value
-      const uint8_t o = cpu_rb(ctx, RB_A),
-                    n = mmu_rb(ctx, cpu_rw(ctx, RW_HL)),
-                    c = FLAG(ctx, C) ? 1 : 0,
-                    v = o - (n + c);
-
-      // write value
-      cpu_wb(ctx, RB_A, v);
-
-      // set flags
-      cpu_wf(ctx, F_Z | F_N | F_H | F_C, (
-        (v ? 0 : F_Z) |
-        F_N |
-        ((((o & 0x0F) - (n & 0x0F)) & 0xF0) ? F_H : 0) |
-        (((((uint16_t) n) + c) > o) ? F_C : 0) // FIXME: is this right for sub/sbc?
-      ));
+      sbc(ctx, mmu_rb(ctx, cpu_rw(ctx, RW_HL)));
     }
 
     static void
-    and_rb(
+    and(
       gb_t * const ctx,
-      const rb_t reg
+      const uint8_t val
     ) {
       // get value
-      const uint8_t v = cpu_rb(ctx, RB_A) & cpu_rb(ctx, reg);
+      const uint8_t v = cpu_rb(ctx, RB_A) & val;
 
       // write value
       cpu_wb(ctx, RB_A, v);
@@ -9570,20 +9528,18 @@ templates:
     }
 
     static void
+    and_rb(
+      gb_t * const ctx,
+      const rb_t reg
+    ) {
+      and(ctx, cpu_rb(ctx, reg));
+    }
+
+    static void
     and_hl_ptr(
       gb_t * const ctx
     ) {
-      // get value
-      const uint8_t v = cpu_rb(ctx, RB_A) & mmu_rb(ctx, cpu_rb(ctx, RW_HL));
-
-      // write value
-      cpu_wb(ctx, RB_A, v);
-
-      // set flags
-      cpu_wf(ctx, F_Z | F_N | F_H | F_C, (
-        (v ? 0 : F_Z) |
-        F_H
-      ));
+      and(ctx, mmu_rb(ctx, cpu_rb(ctx, RW_HL)));
     }
 
     static void
@@ -9591,8 +9547,16 @@ templates:
       gb_t * const ctx,
       const uint16_t addr
     ) {
+      and(ctx, mmu_rb(ctx, addr));
+    }
+
+    static void
+    xor(
+      gb_t * const ctx,
+      const uint8_t val
+    ) {
       // get value
-      const uint8_t v = cpu_rb(ctx, RB_A) & mmu_rb(ctx, addr);
+      const uint8_t v = cpu_rb(ctx, RB_A) ^ val;
 
       // write value
       cpu_wb(ctx, RB_A, v);
@@ -9608,32 +9572,14 @@ templates:
       gb_t * const ctx,
       const rb_t reg
     ) {
-      // get value
-      const uint8_t v = cpu_rb(ctx, RB_A) ^ cpu_rb(ctx, reg);
-
-      // write value
-      cpu_wb(ctx, RB_A, v);
-
-      // set flags
-      cpu_wf(ctx, F_Z | F_N | F_H | F_C, (
-        (v ? 0 : F_Z)
-      ));
+      xor(ctx, cpu_rb(ctx, reg));
     }
 
     static void
     xor_hl_ptr(
       gb_t * const ctx
     ) {
-      // get value
-      const uint8_t v = cpu_rb(ctx, RB_A) ^ mmu_rb(ctx, cpu_rw(ctx, RW_HL));
-
-      // write value
-      cpu_wb(ctx, RB_A, v);
-
-      // set flags
-      cpu_wf(ctx, F_Z | F_N | F_H | F_C, (
-        (v ? 0 : F_Z)
-      ));
+      xor(ctx, mmu_rb(ctx, cpu_rw(ctx, RW_HL)));
     }
 
     static void
@@ -9641,8 +9587,15 @@ templates:
       gb_t * const ctx,
       const uint16_t addr
     ) {
-      // get value
-      const uint8_t v = cpu_rb(ctx, RB_A) ^ mmu_rb(ctx, addr);
+      xor(ctx, mmu_rb(ctx, addr));
+    }
+
+    static void
+    or(
+      gb_t * const ctx,
+      const uint8_t val
+    ) {
+      const uint8_t v = cpu_rb(ctx, RB_A) | val;
 
       // write value
       cpu_wb(ctx, RB_A, v);
@@ -9658,30 +9611,14 @@ templates:
       gb_t * const ctx,
       const rb_t reg
     ) {
-      const uint8_t v = cpu_rb(ctx, RB_A) | cpu_rb(ctx, reg);
-      // write value
-      cpu_wb(ctx, RB_A, v);
-
-      // set flags
-      cpu_wf(ctx, F_Z | F_N | F_H | F_C, (
-        (v ? 0 : F_Z)
-      ));
+      or(ctx, cpu_rb(ctx, reg));
     }
 
     static void
     or_hl_ptr(
       gb_t * const ctx
     ) {
-      // get value
-      const uint8_t v = cpu_rb(ctx, RB_A) | mmu_rb(ctx, cpu_rw(ctx, RW_HL));
-
-      // write value
-      cpu_wb(ctx, RB_A, v);
-
-      // set flags
-      cpu_wf(ctx, F_Z | F_N | F_H | F_C, (
-        (v ? 0 : F_Z)
-      ));
+      or(ctx, mmu_rb(ctx, cpu_rw(ctx, RW_HL)));
     }
 
     static void
@@ -9689,26 +9626,16 @@ templates:
       gb_t * const ctx,
       const uint16_t addr
     ) {
-      // get value
-      const uint8_t v = cpu_rb(ctx, RB_A) | mmu_rb(ctx, addr);
-
-      // write value
-      cpu_wb(ctx, RB_A, v);
-
-      // set flags
-      cpu_wf(ctx, F_Z | F_N | F_H | F_C, (
-        (v ? 0 : F_Z)
-      ));
+      or(ctx, mmu_rb(ctx, addr));
     }
 
     static void
-    cp_rb(
+    cp(
       gb_t * const ctx,
-      const rb_t reg
+      const uint8_t n
     ) {
       // get old and new value
       const uint8_t o = cpu_rb(ctx, RB_A),
-                    n = cpu_rb(ctx, reg),
                     v = o - n;
 
       // set flags
@@ -9721,21 +9648,18 @@ templates:
     }
 
     static void
+    cp_rb(
+      gb_t * const ctx,
+      const rb_t reg
+    ) {
+      cp(ctx, cpu_rb(ctx, reg));
+    }
+
+    static void
     cp_hl_ptr(
       gb_t * const ctx
     ) {
-      // get old and new value
-      const uint8_t o = cpu_rb(ctx, RB_A),
-                    n = mmu_rb(ctx, cpu_rw(ctx, RW_HL)),
-                    v = o - n;
-
-      // set flags
-      cpu_wf(ctx, F_Z | F_N | F_H | F_C, (
-        (v ? 0 : F_Z) |
-        F_N |
-        ((((o & 0x0F) - (n & 0x0F)) & 0xF0) ? F_H : 0) |
-        ((n > o) ? F_C : 0)
-      ));
+      cp(ctx, mmu_rb(ctx, cpu_rw(ctx, RW_HL)));
     }
 
     static void
@@ -9743,18 +9667,7 @@ templates:
       gb_t * const ctx,
       const uint16_t addr
     ) {
-      // get old and new value
-      const uint8_t o = cpu_rb(ctx, RB_A),
-                    n = mmu_rb(ctx, addr),
-                    v = o - n;
-
-      // set flags
-      cpu_wf(ctx, F_Z | F_N | F_H | F_C, (
-        (v ? 0 : F_Z) |
-        F_N |
-        ((((o & 0x0F) - (n & 0x0F)) & 0xF0) ? F_H : 0) |
-        ((n > o) ? F_C : 0)
-      ));
+      cp(ctx, mmu_rb(ctx, addr));
     }
 
     static void
@@ -9787,42 +9700,56 @@ templates:
       cpu_ww(ctx, RW_SP, sp + 2);
     }
 
-    static void
-    rlc_rb(
+    static uint8_t
+    rlc(
       gb_t * const ctx,
-      const rb_t reg
+      const uint8_t o
     ) {
-      // get old value, carry bit, and new value
-      const uint8_t o = cpu_rb(ctx, reg),
-                    c = (o & 0x80) ? 1 : 0,
+      // get carry bit and new value
+      const uint8_t c = (o & 0x80) ? 1 : 0,
                     n = ((o & 0x7F) << 1) + c;
-      // set value
-      cpu_wb(ctx, reg, n);
-
       // set flags
       cpu_wf(ctx, F_Z | F_N | F_H | F_C, (
         (n ? 0 : F_Z) |
         (c ? F_C : 0)
       ));
+
+      // return new value
+      return n;
+    }
+
+    static void
+    rlc_rb(
+      gb_t * const ctx,
+      const rb_t reg
+    ) {
+      cpu_wb(ctx, reg, rlc(ctx, cpu_rb(ctx, reg)));
     }
 
     static void
     rlc_hl(
       gb_t * const ctx
     ) {
-      // get old value, carry bit, and new value
       const uint16_t addr = cpu_rw(ctx, RW_HL);
-      const uint8_t o = mmu_rb(ctx, addr),
-                    c = (o & 0x80) ? 1 : 0,
-                    n = ((o & 0x7F) << 1) + c;
-      // set value
-      mmu_wb(ctx, addr, n);
+      mmu_wb(ctx, addr, rlc(ctx, mmu_rb(ctx, addr)));
+    }
 
+    static uint8_t
+    rrc(
+      gb_t * const ctx,
+      const uint8_t o
+    ) {
+      // get  carry bit and new value
+      const uint8_t c = (o & 0x01) ? 1 : 0,
+                    n = (o >> 1) | (c ? 0x80 : 0);
       // set flags
       cpu_wf(ctx, F_Z | F_N | F_H | F_C, (
         (n ? 0 : F_Z) |
         (c ? F_C : 0)
       ));
+
+      // return new value
+      return n;
     }
 
     static void
@@ -9830,37 +9757,32 @@ templates:
       gb_t * const ctx,
       const rb_t reg
     ) {
-      // get old value, carry bit, and new value
-      const uint8_t o = cpu_rb(ctx, reg),
-                    c = (o & 0x01) ? 1 : 0,
-                    n = (o >> 1) | (c ? 0x80 : 0);
-      // set value
-      cpu_wb(ctx, reg, n);
-
-      // set flags
-      cpu_wf(ctx, F_Z | F_N | F_H | F_C, (
-        (n ? 0 : F_Z) |
-        (c ? F_C : 0)
-      ));
+      cpu_wb(ctx, reg, rrc(ctx, cpu_rb(ctx, reg)));
     }
 
     static void
     rrc_hl_ptr(
       gb_t * const ctx
     ) {
-      // get old value, carry bit, and new value
       const uint16_t addr = cpu_rw(ctx, RW_HL);
-      const uint8_t o = mmu_rb(ctx, addr),
-                    c = (o & 0x01) ? 1 : 0,
-                    n = (o >> 1) | (c ? 0x80 : 0);
-      // set value
-      mmu_wb(ctx, addr, n);
+      mmu_wb(ctx, addr, rrc(ctx, mmu_rb(ctx, addr)));
+    }
+
+    static uint8_t
+    rl(
+      gb_t * const ctx,
+      const uint8_t o
+    ) {
+      const uint8_t n = (o << 1) + (FLAG(ctx, C) ? 1 : 0);
 
       // set flags
       cpu_wf(ctx, F_Z | F_N | F_H | F_C, (
         (n ? 0 : F_Z) |
-        (c ? F_C : 0)
+        ((o & 0x80) ? F_C : 0)
       ));
+
+      // return result
+      return n;
     }
 
     static void
@@ -9868,35 +9790,32 @@ templates:
       gb_t * const ctx,
       const rb_t reg
     ) {
-      // get old value and new value
-      const uint8_t o = cpu_rb(ctx, reg),
-                    n = (o << 1) + (FLAG(ctx, C) ? 1 : 0);
-      // set value
-      cpu_wb(ctx, reg, n);
-
-      // set flags
-      cpu_wf(ctx, F_Z | F_N | F_H | F_C, (
-        (n ? 0 : F_Z) |
-        ((o & 0x80) ? F_C : 0)
-      ));
+      cpu_wb(ctx, reg, rl(ctx, cpu_rb(ctx, reg)));
     }
 
     static void
     rl_hl_ptr(
       gb_t * const ctx
     ) {
-      // get old value and new value
       const uint16_t addr = cpu_rw(ctx, RW_HL);
-      const uint8_t o = mmu_rb(ctx, addr),
-                    n = (o << 1) + (FLAG(ctx, C) ? 1 : 0);
-      // set value
-      mmu_wb(ctx, addr, n);
+      mmu_wb(ctx, addr, rl(ctx, mmu_rb(ctx, addr)));
+    }
+
+    static uint8_t
+    rr(
+      gb_t * const ctx,
+      const uint8_t o
+    ) {
+      const uint8_t n = (FLAG(ctx, C) ? 0x80 : 0) | (o >> 1);
 
       // set flags
       cpu_wf(ctx, F_Z | F_N | F_H | F_C, (
         (n ? 0 : F_Z) |
-        ((o & 0x80) ? F_C : 0)
+        ((o & 0x01) ? F_C : 0)
       ));
+
+      // return result
+      return n;
     }
 
     static void
@@ -9904,35 +9823,32 @@ templates:
       gb_t * const ctx,
       const rb_t reg
     ) {
-      // get old value and new value
-      const uint8_t o = cpu_rb(ctx, reg),
-                    n = (FLAG(ctx, C) ? 0x80 : 0) | (o >> 1);
-      // set value
-      cpu_wb(ctx, reg, n);
-
-      // set flags
-      cpu_wf(ctx, F_Z | F_N | F_H | F_C, (
-        (n ? 0 : F_Z) |
-        ((o & 0x01) ? F_C : 0)
-      ));
+      cpu_wb(ctx, reg, rr(ctx, cpu_rb(ctx, reg)));
     }
 
     static void
     rr_hl_ptr(
       gb_t * const ctx
     ) {
-      // get old value and new value
       const uint16_t addr = cpu_rw(ctx, RW_HL);
-      const uint8_t o = mmu_rb(ctx, addr),
-                    n = (FLAG(ctx, C) ? 0x80 : 0) | (o >> 1);
-      // set value
-      mmu_wb(ctx, addr, n);
+      mmu_wb(ctx, addr, rr(ctx, mmu_rb(ctx, addr)));
+    }
+
+    static uint8_t
+    sla(
+      gb_t * const ctx,
+      const uint8_t o
+    ) {
+      const uint8_t v = o << 1;
 
       // set flags
       cpu_wf(ctx, F_Z | F_N | F_H | F_C, (
-        (n ? 0 : F_Z) |
-        ((o & 0x01) ? F_C : 0)
+        (v ? 0 : F_Z) |
+        ((o & 0x80) ? F_C : 0)
       ));
+
+      // return value
+      return v;
     }
 
     static void
@@ -9940,17 +9856,7 @@ templates:
       gb_t * const ctx,
       const rb_t reg
     ) {
-      const uint8_t o = cpu_rb(ctx, reg),
-                    v = o << 1;
-
-      // set value
-      cpu_wb(ctx, reg, v);
-
-      // set flags
-      cpu_wf(ctx, F_Z | F_N | F_H | F_C, (
-        (v ? 0 : F_Z) |
-        ((o & 0x80) ? F_C : 0)
-      ));
+      cpu_wb(ctx, reg, sla(ctx, cpu_rb(ctx, reg)));
     }
 
     static void
@@ -9958,17 +9864,24 @@ templates:
       gb_t * const ctx
     ) {
       const uint16_t addr = cpu_rw(ctx, RW_HL);
-      const uint8_t o = mmu_rb(ctx, addr),
-                    v = o << 1;
+      mmu_wb(ctx, addr, sla(ctx, mmu_rb(ctx, addr)));
+    }
 
-      // set value
-      mmu_wb(ctx, addr, v);
+    static uint8_t
+    sra(
+      gb_t * const ctx,
+      const uint8_t o
+    ) {
+      const uint8_t v = ((o & 0x80) ? 0x80 : 0) | (o >> 1);
 
       // set flags
       cpu_wf(ctx, F_Z | F_N | F_H | F_C, (
         (v ? 0 : F_Z) |
-        ((o & 0x80) ? F_C : 0)
+        ((o & 0x01) ? F_C : 0)
       ));
+
+      // return result
+      return v;
     }
 
     static void
@@ -9976,17 +9889,7 @@ templates:
       gb_t * const ctx,
       const rb_t reg
     ) {
-      const uint8_t o = cpu_rb(ctx, reg),
-                    v = ((o & 0x80) ? 0x80 : 0) | (o >> 1);
-
-      // set value
-      cpu_wb(ctx, reg, v);
-
-      // set flags
-      cpu_wf(ctx, F_Z | F_N | F_H | F_C, (
-        (v ? 0 : F_Z) |
-        ((o & 0x01) ? F_C : 0)
-      ));
+      cpu_wb(ctx, reg, sra(ctx, cpu_rb(ctx, reg)));
     }
 
     static void
@@ -9994,17 +9897,24 @@ templates:
       gb_t * const ctx
     ) {
       const uint16_t addr = cpu_rw(ctx, RW_HL);
-      const uint8_t o = mmu_rb(ctx, addr),
-                    v = ((o & 0x80) ? 0x80 : 0) | (o >> 1);
+      mmu_wb(ctx, addr, sra(ctx, mmu_rb(ctx, addr)));
+    }
 
-      // set value
-      mmu_wb(ctx, addr, v);
+    static uint8_t
+    srl(
+      gb_t * const ctx,
+      const uint8_t o
+    ) {
+      const uint8_t v = (o >> 1);
 
       // set flags
       cpu_wf(ctx, F_Z | F_N | F_H | F_C, (
         (v ? 0 : F_Z) |
         ((o & 0x01) ? F_C : 0)
       ));
+
+      // return result
+      return v;
     }
 
     static void
@@ -10012,17 +9922,7 @@ templates:
       gb_t * const ctx,
       const rb_t reg
     ) {
-      const uint8_t o = cpu_rb(ctx, reg),
-                    v = (o >> 1);
-
-      // set value
-      cpu_wb(ctx, reg, v);
-
-      // set flags
-      cpu_wf(ctx, F_Z | F_N | F_H | F_C, (
-        (v ? 0 : F_Z) |
-        ((o & 0x01) ? F_C : 0)
-      ));
+      cpu_wb(ctx, reg, srl(ctx, cpu_rb(ctx, reg)));
     }
 
     static void
@@ -10030,17 +9930,7 @@ templates:
       gb_t * const ctx
     ) {
       const uint16_t addr = cpu_rw(ctx, RW_HL);
-      const uint8_t o = mmu_rb(ctx, addr),
-                    v = (o >> 1);
-
-      // set value
-      mmu_wb(ctx, addr, v);
-
-      // set flags
-      cpu_wf(ctx, F_Z | F_N | F_H | F_C, (
-        (v ? 0 : F_Z) |
-        ((o & 0x01) ? F_C : 0)
-      ));
+      mmu_wb(ctx, addr, srl(ctx, mmu_rb(ctx, addr)));
     }
 
     static void
@@ -10121,21 +10011,25 @@ templates:
       mmu_wb(ctx, addr, mmu_rb(ctx, addr) | ~(bit ? (1 << bit) : 1));
     }
 
-    static void
-    swap_rb(
+    static uint8_t
+    swap(
       gb_t * const ctx,
-      const rb_t reg
+      const uint8_t v
     ) {
-      // get old value
-      const uint8_t v = cpu_rb(ctx, reg);
-
-      // write value
-      cpu_wb(ctx, reg, ((v & (0x0F)) << 4) | ((v & 0xF0) >> 4));
-
       // write flags
       cpu_wf(ctx, F_Z | F_N | F_H | F_C, (
         (v ? 0 : F_Z)
       ));
+
+      return ((v & (0x0F)) << 4) | ((v & 0xF0) >> 4);
+    }
+
+    static void
+    swap_rb(
+      gb_t * const ctx,
+      const rb_t reg
+    ) {
+      cpu_wb(ctx, reg, swap(ctx, cpu_rb(ctx, reg)));
     }
 
     static void
@@ -10143,15 +10037,7 @@ templates:
       gb_t * const ctx
     ) {
       const uint16_t addr = cpu_rw(ctx, RW_HL);
-      const uint8_t v = mmu_rb(ctx, addr);
-
-      // write value
-      mmu_wb(ctx, addr, ((v & (0x0F)) << 4) | ((v & 0xF0) >> 4));
-
-      // write flags
-      cpu_wf(ctx, F_Z | F_N | F_H | F_C, (
-        (v ? 0 : F_Z)
-      ));
+      mmu_wb(ctx, addr, swap(ctx, mmu_rb(ctx, addr)));
     }
 
     static void
@@ -10203,6 +10089,13 @@ templates:
       }
     }
 
+    static uint8_t
+    gpu_get_mode(
+      const gb_t * const ctx
+    ) {
+      return ctx->gpu.mode & 0x3;
+    }
+
     static void
     gpu_set_mode(
       gb_t * const ctx,
@@ -10210,18 +10103,29 @@ templates:
     ) {
       ctx->gpu.mode = mode;
 
-      switch (mode) {
+      if (ctx->config && ctx->config->on_gpu_set_mode) {
+        ctx->config->on_gpu_set_mode(ctx, mode);
+      }
+
+      switch (gpu_get_mode(ctx)) {
       case GPU_MODE_OAM:
         if (ctx->gpu.stat & (1 << 5)) {
-          // vblank interrupt is enabled, trigger LCDC interrupt
+          // lcd status interrupt is enabled, trigger LCDC interrupt
           mmu_wb(ctx, PORT_IF, mmu_rb(ctx, PORT_IF) & (1 << 1));
         }
 
         break;
+      case GPU_MODE_VRAM:
+        // do nothing
+        break;
       case GPU_MODE_HBLANK:
         if (ctx->gpu.stat & (1 << 3)) {
-          // hblank interrupt is enabled, trigger LCDC interrupt
+          // lcd status interrupt is enabled, trigger LCDC interrupt
           mmu_wb(ctx, PORT_IF, mmu_rb(ctx, PORT_IF) & (1 << 1));
+
+          if (ctx->config && ctx->config->on_hblank) {
+            ctx->config->on_hblank(ctx);
+          }
         }
 
         break;
@@ -10229,12 +10133,13 @@ templates:
         if (ctx->gpu.stat & (1 << 4)) {
           // vblank interrupt is enabled, trigger vblank interrupt
           mmu_wb(ctx, PORT_IF, mmu_rb(ctx, PORT_IF) & 1);
+
+          if (ctx->config && ctx->config->on_vblank) {
+            ctx->config->on_vblank(ctx);
+          }
         }
 
         break;
-      default:
-        // do nothing
-        break;
       }
     }
 
@@ -10332,19 +10237,21 @@ templates:
       // increment clock
       ctx->gpu.clock += clock;
 
-      switch (ctx->gpu.mode) {
+      if (ctx->config && ctx->config->on_gpu_step) {
+        ctx->config->on_gpu_step(ctx);
+      }
+
+      switch (gpu_get_mode(ctx)) {
       case GPU_MODE_OAM:
-        if (ctx->gpu.clock > 80) {
+        if (ctx->gpu.clock > 79) {
           // clear clock, set mode
           ctx->gpu.clock = 0;
-          ctx->gpu.mode = GPU_MODE_VRAM;
+          gpu_set_mode(ctx, GPU_MODE_VRAM);
         }
 
         break;
       case GPU_MODE_VRAM:
-        if (ctx->gpu.clock > 172) {
-          // TODO: write scanline
-
+        if (ctx->gpu.clock > 171) {
           // clear clock, set mode
           ctx->gpu.clock = 0;
           gpu_set_mode(ctx, GPU_MODE_HBLANK);
@@ -10352,13 +10259,13 @@ templates:
 
         break;
       case GPU_MODE_HBLANK:
-        if (ctx->gpu.clock > 204) {
+        if (ctx->gpu.clock > 203) {
           // clear clock, draw line, increment line
           ctx->gpu.clock = 0;
           gpu_draw(ctx);
           gpu_set_line(ctx, ctx->gpu.line + 1);
 
-          if (ctx->gpu.line < 143) {
+          if (ctx->gpu.line < 144) {
             // set mode
             gpu_set_mode(ctx, GPU_MODE_OAM);
           } else {
@@ -10386,9 +10293,6 @@ templates:
         }
 
         break;
-      default:
-        /* do nothing, unknown gpu mode */
-        break;
       }
     }
 
@@ -10412,6 +10316,11 @@ templates:
 
           // trigger timer interrupt
           mmu_wb(ctx, PORT_IF, mmu_rb(ctx, PORT_IF) & (1 << 2));
+
+          if (ctx->config && ctx->config->on_timer) {
+            // notify callback
+            ctx->config->on_timer(ctx);
+          }
         }
       }
     }
@@ -10450,8 +10359,8 @@ templates:
         // return 4 clock step
         return 4;
       } else {
-        // return 4 clock step
-        return 0;
+        // return 0 clock step
+        return 1;
       }
     }
 
@@ -10461,7 +10370,7 @@ templates:
     ) {
       if (ctx->cpu.state == GB_CPU_STATE_RUN) {
         if (ctx->cpu.ime) {
-          // get interrupt vector (masked against enabled interrupts)
+          // get interrupt flags, masked against enabled interrupts
           const uint8_t iv = mmu_rb(ctx, PORT_IE) & mmu_rb(ctx, PORT_IF);
           bool done = false;
 
@@ -10546,6 +10455,59 @@ templates:
       gpu_step(ctx, clock);
     }
 
+    static const struct {
+      const char * const name;
+      const uint8_t len;
+    } OPS[] = {
+    <%= ops %>
+    };
+
+    bool
+    gb_disasm(
+      gb_t * const ctx,
+      char * const dst_buf,
+      size_t * const dst_len
+    ) {
+      if (!dst_buf || !dst_len) {
+        return false;
+      }
+
+      const uint8_t op = mmu_rb(ctx, cpu_rw(ctx, RW_PC));
+      const uint16_t ofs = (op == 0xCB) ? (255 + mmu_rb(ctx, cpu_rw(ctx, RW_PC) + 1)) : op;
+      int len = -1;
+
+      switch (OPS[ofs].len) {
+      case 1:
+        len = snprintf(dst_buf, *dst_len, "%s; %02x",
+          OPS[ofs].name,
+          op
+        );
+
+        break;
+      case 2:
+        len = snprintf(dst_buf, *dst_len, "%s; %02x %02x",
+          OPS[ofs].name,
+          op,
+          mmu_rb(ctx, cpu_rw(ctx, RW_PC) + 1)
+        );
+
+        break;
+      case 3:
+        len = snprintf(dst_buf, *dst_len, "%s; %02x %02x %02x",
+          OPS[ofs].name,
+          op,
+          mmu_rb(ctx, cpu_rw(ctx, RW_PC) + 1),
+          mmu_rb(ctx, cpu_rw(ctx, RW_PC) + 2)
+        );
+
+        break;
+      default:
+        len = snprintf(dst_buf, *dst_len, "%s", OPS[ofs].name);
+      }
+
+      return (len > 0) && (len < (int) *dst_len);
+    }
+
     void
     gb_set_buttons(
       gb_t * const ctx,
@@ -10584,9 +10546,6 @@ templates:
     cpu_init(
       gb_t * const ctx
     ) {
-      // set cpu state
-      cpu_set_state(ctx, 0, GB_CPU_STATE_RUN);
-
       // init cpu registers
       // (src: TCAGBD.pdf, 3.2)
       // TODO: this varies by model (DGB/CGB/SGB/GBA/etc)
@@ -10596,16 +10555,23 @@ templates:
       ctx->cpu.rs[RW_HL] = 0x014D;
       ctx->cpu.rs[RW_SP] = 0xFFFE;
       ctx->cpu.rs[RW_PC] = 0x0100;
+
+      // set cpu state
+      cpu_set_state(ctx, 0, GB_CPU_STATE_RUN);
     }
 
     void gb_init(
       gb_t * const ctx,
-      const uint8_t * const rom,
+      const gb_config_t * const config,
+      const uint8_t *rom,
       const uint32_t rom_size
     ) {
       // clear context
       memset(ctx, 0, sizeof(gb_t));
 
+      // save config
+      ctx->config = config;
+
       // init cpu
       cpu_init(ctx);
 
@@ -10614,6 +10580,6 @@ templates:
       ctx->mmu.rom_size = rom_size;
 
       // init gpu
-      ctx->gpu.mode = GPU_MODE_OAM;
-      ctx->gpu.line = 0;
+      gpu_set_mode(ctx, GPU_MODE_OAM);
+      gpu_set_line(ctx, 0);
     }