chore: resolve bulid error after rebase on main

use memset like most other allocations in this emu

src/core/Bus.zig

@@ -12,7 +12,7 @@ const Apu = @import("apu.zig").Apu;
 const DmaTuple = @import("bus/dma.zig").DmaTuple;
 const TimerTuple = @import("bus/timer.zig").TimerTuple;
 const Scheduler = @import("scheduler.zig").Scheduler;
-const FilePaths = @import("util.zig").FilePaths;
+const FilePaths = @import("../util.zig").FilePaths;
 
 const io = @import("bus/io.zig");
 const Allocator = std.mem.Allocator;
@@ -20,7 +20,7 @@ const log = std.log.scoped(.Bus);
 
 const createDmaTuple = @import("bus/dma.zig").create;
 const createTimerTuple = @import("bus/timer.zig").create;
-const rotr = @import("util.zig").rotr;
+const rotr = @import("../util.zig").rotr;
 
 const timings: [2][0x10]u8 = [_][0x10]u8{
     // BIOS, Unused, EWRAM, IWRAM, I/0, PALRAM, VRAM, OAM, ROM0, ROM0, ROM1, ROM1, ROM2, ROM2, SRAM, Unused

src/core/apu.zig

@@ -1,7 +1,7 @@
 const std = @import("std");
 const SDL = @import("sdl2");
 const io = @import("bus/io.zig");
-const util = @import("util.zig");
+const util = @import("../util.zig");
 
 const Arm7tdmi = @import("cpu.zig").Arm7tdmi;
 const Scheduler = @import("scheduler.zig").Scheduler;
@@ -9,7 +9,7 @@ const Scheduler = @import("scheduler.zig").Scheduler;
 const SoundFifo = std.fifo.LinearFifo(u8, .{ .Static = 0x20 });
 const AudioDeviceId = SDL.SDL_AudioDeviceID;
 
-const intToBytes = @import("util.zig").intToBytes;
+const intToBytes = @import("../util.zig").intToBytes;
 const log = std.log.scoped(.APU);
 
 pub const host_sample_rate = 1 << 15;

src/core/bus/backup.zig

@@ -2,8 +2,8 @@ const std = @import("std");
 const Allocator = std.mem.Allocator;
 const log = std.log.scoped(.Backup);
 
-const escape = @import("../util.zig").escape;
-const span = @import("../util.zig").span;
+const escape = @import("../../util.zig").escape;
+const span = @import("../../util.zig").span;
 
 const backup_kinds = [6]Needle{
     .{ .str = "EEPROM_V", .kind = .Eeprom },

src/core/bus/dma.zig

@@ -1,5 +1,5 @@
 const std = @import("std");
-const util = @import("../util.zig");
+const util = @import("../../util.zig");
 
 const DmaControl = @import("io.zig").DmaControl;
 const Bus = @import("../Bus.zig");
@@ -227,7 +227,7 @@ fn DmaController(comptime id: u2) type {
             }
         }
 
-        pub fn pollBlankingDma(self: *Self, comptime kind: DmaKind) void {
+        pub fn pollBlanking(self: *Self, comptime kind: DmaKind) void {
             if (self.in_progress) return; // If there's an ongoing DMA Transfer, exit early
 
             // No ongoing DMA Transfer, We want to check if we should repeat an existing one
@@ -271,11 +271,11 @@ fn DmaController(comptime id: u2) type {
     };
 }
 
-pub fn pollBlankingDma(bus: *Bus, comptime kind: DmaKind) void {
-    bus.dma[0].pollBlankingDma(kind);
-    bus.dma[1].pollBlankingDma(kind);
-    bus.dma[2].pollBlankingDma(kind);
-    bus.dma[3].pollBlankingDma(kind);
+pub fn onBlanking(bus: *Bus, comptime kind: DmaKind) void {
+    bus.dma[0].pollBlanking(kind);
+    bus.dma[1].pollBlanking(kind);
+    bus.dma[2].pollBlanking(kind);
+    bus.dma[3].pollBlanking(kind);
 }
 
 const Adjustment = enum(u2) {

src/core/bus/io.zig

@@ -3,7 +3,7 @@ const builtin = @import("builtin");
 const timer = @import("timer.zig");
 const dma = @import("dma.zig");
 const apu = @import("../apu.zig");
-const util = @import("../util.zig");
+const util = @import("../../util.zig");
 
 const Bit = @import("bitfield").Bit;
 const Bitfield = @import("bitfield").Bitfield;
@@ -302,6 +302,14 @@ pub fn write(bus: *Bus, comptime T: type, address: u32, value: T) void {
             0x0400_0009 => bus.ppu.bg[0].cnt.raw = (@as(u16, value) << 8) | (bus.ppu.bg[0].cnt.raw & 0xFF),
             0x0400_000A => bus.ppu.bg[1].cnt.raw = (bus.ppu.bg[1].cnt.raw & 0xFF00) | value,
             0x0400_000B => bus.ppu.bg[1].cnt.raw = (@as(u16, value) << 8) | (bus.ppu.bg[1].cnt.raw & 0xFF),
+            0x0400_0040 => bus.ppu.win.h[0].set(.Lo, value),
+            0x0400_0041 => bus.ppu.win.h[0].set(.Hi, value),
+            0x0400_0042 => bus.ppu.win.h[1].set(.Lo, value),
+            0x0400_0043 => bus.ppu.win.h[1].set(.Hi, value),
+            0x0400_0044 => bus.ppu.win.v[0].set(.Lo, value),
+            0x0400_0045 => bus.ppu.win.v[0].set(.Hi, value),
+            0x0400_0046 => bus.ppu.win.v[1].set(.Lo, value),
+            0x0400_0047 => bus.ppu.win.v[1].set(.Hi, value),
             0x0400_0048 => bus.ppu.win.setInL(value),
             0x0400_0049 => bus.ppu.win.setInH(value),
             0x0400_004A => bus.ppu.win.setOutL(value),
@@ -459,37 +467,57 @@ pub const BldY = extern union {
     raw: u16,
 };
 
+const u8WriteKind = enum { Hi, Lo };
+
 /// Write-only
 pub const WinH = extern union {
+    const Self = @This();
+
     x2: Bitfield(u16, 0, 8),
     x1: Bitfield(u16, 8, 8),
     raw: u16,
+
+    pub fn set(self: *Self, comptime K: u8WriteKind, value: u8) void {
+        self.raw = switch (K) {
+            .Hi => (@as(u16, value) << 8) | self.raw & 0xFF,
+            .Lo => (self.raw & 0xFF00) | value,
+        };
+    }
 };
 
 /// Write-only
 pub const WinV = extern union {
+    const Self = @This();
+
     y2: Bitfield(u16, 0, 8),
     y1: Bitfield(u16, 8, 8),
     raw: u16,
+
+    pub fn set(self: *Self, comptime K: u8WriteKind, value: u8) void {
+        self.raw = switch (K) {
+            .Hi => (@as(u16, value) << 8) | self.raw & 0xFF,
+            .Lo => (self.raw & 0xFF00) | value,
+        };
+    }
 };
 
 pub const WinIn = extern union {
     w0_bg: Bitfield(u16, 0, 4),
     w0_obj: Bit(u16, 4),
-    w0_colour: Bit(u16, 5),
+    w0_bld: Bit(u16, 5),
     w1_bg: Bitfield(u16, 8, 4),
     w1_obj: Bit(u16, 12),
-    w1_colour: Bit(u16, 13),
+    w1_bld: Bit(u16, 13),
     raw: u16,
 };
 
 pub const WinOut = extern union {
     out_bg: Bitfield(u16, 0, 4),
     out_obj: Bit(u16, 4),
-    out_colour: Bit(u16, 5),
+    out_bld: Bit(u16, 5),
     obj_bg: Bitfield(u16, 8, 4),
     obj_obj: Bit(u16, 12),
-    obj_colour: Bit(u16, 13),
+    obj_bld: Bit(u16, 13),
     raw: u16,
 };
 

src/core/bus/timer.zig

@@ -1,5 +1,5 @@
 const std = @import("std");
-const util = @import("../util.zig");
+const util = @import("../../util.zig");
 
 const TimerControl = @import("io.zig").TimerControl;
 const Io = @import("io.zig").Io;

src/core/cpu.zig

@@ -1,12 +1,12 @@
 const std = @import("std");
-const util = @import("util.zig");
+const util = @import("../util.zig");
 
 const Bus = @import("Bus.zig");
 const Bit = @import("bitfield").Bit;
 const Bitfield = @import("bitfield").Bitfield;
 const Scheduler = @import("scheduler.zig").Scheduler;
-const FilePaths = @import("util.zig").FilePaths;
-const Logger = @import("util.zig").Logger;
+const FilePaths = @import("../util.zig").FilePaths;
+const Logger = @import("../util.zig").Logger;
 
 const File = std.fs.File;
 
@@ -125,7 +125,7 @@ pub const thumb = struct {
     const swi = @import("cpu/thumb/software_interrupt.zig").fmt17;
     const branch = @import("cpu/thumb/branch.zig");
 
-    /// Determine index into THUMB InstrFn LUT 
+    /// Determine index into THUMB InstrFn LUT
     fn idx(opcode: u16) u10 {
         return @truncate(u10, opcode >> 6);
     }

src/core/cpu/arm/branch.zig

@@ -4,7 +4,7 @@ const Bus = @import("../../Bus.zig");
 const Arm7tdmi = @import("../../cpu.zig").Arm7tdmi;
 const InstrFn = @import("../../cpu.zig").arm.InstrFn;
 
-const sext = @import("../../util.zig").sext;
+const sext = @import("../../../util.zig").sext;
 
 pub fn branch(comptime L: bool) InstrFn {
     return struct {

src/core/cpu/arm/half_signed_data_transfer.zig

@@ -4,8 +4,8 @@ const Bus = @import("../../Bus.zig");
 const Arm7tdmi = @import("../../cpu.zig").Arm7tdmi;
 const InstrFn = @import("../../cpu.zig").arm.InstrFn;
 
-const sext = @import("../../util.zig").sext;
-const rotr = @import("../../util.zig").rotr;
+const sext = @import("../../../util.zig").sext;
+const rotr = @import("../../../util.zig").rotr;
 
 pub fn halfAndSignedDataTransfer(comptime P: bool, comptime U: bool, comptime I: bool, comptime W: bool, comptime L: bool) InstrFn {
     return struct {

src/core/cpu/arm/psr_transfer.zig

@@ -7,7 +7,7 @@ const PSR = @import("../../cpu.zig").PSR;
 
 const log = std.log.scoped(.PsrTransfer);
 
-const rotr = @import("../../util.zig").rotr;
+const rotr = @import("../../../util.zig").rotr;
 
 pub fn psrTransfer(comptime I: bool, comptime R: bool, comptime kind: u2) InstrFn {
     return struct {

src/core/cpu/arm/single_data_swap.zig

@@ -4,7 +4,7 @@ const Bus = @import("../../Bus.zig");
 const Arm7tdmi = @import("../../cpu.zig").Arm7tdmi;
 const InstrFn = @import("../../cpu.zig").arm.InstrFn;
 
-const rotr = @import("../../util.zig").rotr;
+const rotr = @import("../../../util.zig").rotr;
 
 pub fn singleDataSwap(comptime B: bool) InstrFn {
     return struct {

src/core/cpu/arm/single_data_transfer.zig

@@ -1,12 +1,12 @@
 const std = @import("std");
-const util = @import("../../util.zig");
+const util = @import("../../../util.zig");
 
 const shifter = @import("../barrel_shifter.zig");
 const Bus = @import("../../Bus.zig");
 const Arm7tdmi = @import("../../cpu.zig").Arm7tdmi;
 const InstrFn = @import("../../cpu.zig").arm.InstrFn;
 
-const rotr = @import("../../util.zig").rotr;
+const rotr = @import("../../../util.zig").rotr;
 
 pub fn singleDataTransfer(comptime I: bool, comptime P: bool, comptime U: bool, comptime B: bool, comptime W: bool, comptime L: bool) InstrFn {
     return struct {

src/core/cpu/barrel_shifter.zig

@@ -3,7 +3,7 @@ const std = @import("std");
 const Arm7tdmi = @import("../cpu.zig").Arm7tdmi;
 const CPSR = @import("../cpu.zig").PSR;
 
-const rotr = @import("../util.zig").rotr;
+const rotr = @import("../../util.zig").rotr;
 
 pub fn execute(comptime S: bool, cpu: *Arm7tdmi, opcode: u32) u32 {
     var result: u32 = undefined;

src/core/cpu/thumb/branch.zig

@@ -3,7 +3,7 @@ const Arm7tdmi = @import("../../cpu.zig").Arm7tdmi;
 const InstrFn = @import("../../cpu.zig").thumb.InstrFn;
 
 const checkCond = @import("../../cpu.zig").checkCond;
-const sext = @import("../../util.zig").sext;
+const sext = @import("../../../util.zig").sext;
 
 pub fn fmt16(comptime cond: u4) InstrFn {
     return struct {

src/core/cpu/thumb/data_transfer.zig

@@ -4,7 +4,8 @@ const Bus = @import("../../Bus.zig");
 const Arm7tdmi = @import("../../cpu.zig").Arm7tdmi;
 const InstrFn = @import("../../cpu.zig").thumb.InstrFn;
 
-const rotr = @import("../../util.zig").rotr;
+const rotr = @import("../../../util.zig").rotr;
+const sext = @import("../../../util.zig").sext;
 
 pub fn fmt6(comptime rd: u3) InstrFn {
     return struct {
@@ -16,8 +17,6 @@ pub fn fmt6(comptime rd: u3) InstrFn {
     }.inner;
 }
 
-const sext = @import("../../util.zig").sext;
-
 pub fn fmt78(comptime op: u2, comptime T: bool) InstrFn {
     return struct {
         fn inner(cpu: *Arm7tdmi, bus: *Bus, opcode: u16) void {

src/core/emu.zig

@@ -4,8 +4,8 @@ const SDL = @import("sdl2");
 const Bus = @import("Bus.zig");
 const Scheduler = @import("scheduler.zig").Scheduler;
 const Arm7tdmi = @import("cpu.zig").Arm7tdmi;
-const FpsTracker = @import("util.zig").FpsTracker;
-const FilePaths = @import("util.zig").FilePaths;
+const FpsTracker = @import("../util.zig").FpsTracker;
+const FilePaths = @import("../util.zig").FilePaths;
 
 const Timer = std.time.Timer;
 const Thread = std.Thread;

src/core/ppu.zig

@@ -1,16 +1,19 @@
 const std = @import("std");
 const io = @import("bus/io.zig");
+const Bit = @import("bitfield").Bit;
+const Bitfield = @import("bitfield").Bitfield;
+const dma = @import("bus/dma.zig");
 
+const Oam = @import("ppu/Oam.zig");
+const Palette = @import("ppu/Palette.zig");
+const Vram = @import("ppu/Vram.zig");
 const EventKind = @import("scheduler.zig").EventKind;
 const Scheduler = @import("scheduler.zig").Scheduler;
 const Arm7tdmi = @import("cpu.zig").Arm7tdmi;
-
-const Bit = @import("bitfield").Bit;
-const Bitfield = @import("bitfield").Bitfield;
+const FrameBuffer = @import("../util.zig").FrameBuffer;
 
 const Allocator = std.mem.Allocator;
-const log = std.log.scoped(.PPU);
-const pollBlankingDma = @import("bus/dma.zig").pollBlankingDma;
+const log = std.log.scoped(.Ppu);
 
 /// This is used to generate byuu / Talurabi's Color Correction algorithm
 const COLOUR_LUT = genColourLut();
@@ -51,14 +54,14 @@ pub const Ppu = struct {
         sched.push(.Draw, 240 * 4);
 
         const sprites = try allocator.create([128]?Sprite);
-        sprites.* = [_]?Sprite{null} ** 128;
+        std.mem.set(?Sprite, sprites, null);
 
         return Self{
             .vram = try Vram.init(allocator),
             .palette = try Palette.init(allocator),
             .oam = try Oam.init(allocator),
             .sched = sched,
-            .framebuf = try FrameBuffer.init(allocator),
+            .framebuf = try FrameBuffer.init(allocator, framebuf_pitch * height),
             .allocator = allocator,
 
             // Registers
@@ -277,16 +280,17 @@ pub const Ppu = struct {
             aff_x += self.aff_bg[n - 2].pa;
             aff_y += self.aff_bg[n - 2].pc;
 
-            if (!shouldDrawBackground(n, self.bldcnt, &self.scanline, i)) continue;
+            const x = @bitCast(u32, ix);
+            const y = @bitCast(u32, iy);
+
+            const win_bounds = self.windowBounds(@truncate(u9, x), @truncate(u8, y));
+            if (!shouldDrawBackground(self, n, win_bounds, i)) continue;
 
             if (self.bg[n].cnt.display_overflow.read()) {
                 ix = if (ix > px_width) @rem(ix, px_width) else if (ix < 0) px_width + @rem(ix, px_width) else ix;
                 iy = if (iy > px_height) @rem(iy, px_height) else if (iy < 0) px_height + @rem(iy, px_height) else iy;
             } else if (ix > px_width or iy > px_height or ix < 0 or iy < 0) continue;
 
-            const x = @bitCast(u32, ix);
-            const y = @bitCast(u32, iy);
-
             const tile_id: u32 = self.vram.read(u8, screen_base + ((y / 8) * @bitCast(u32, tile_width) + (x / 8)));
             const row = y & 7;
             const col = x & 7;
@@ -296,7 +300,7 @@ pub const Ppu = struct {
 
             if (pal_id != 0) {
                 const bgr555 = self.palette.read(u16, pal_id * 2);
-                copyToBackgroundBuffer(n, self.bldcnt, &self.scanline, i, bgr555);
+                self.copyToBackgroundBuffer(n, win_bounds, i, bgr555);
             }
         }
 
@@ -305,7 +309,7 @@ pub const Ppu = struct {
         self.aff_bg[n - 2].y_latch.? += self.aff_bg[n - 2].pd; // PD is added to BGxY
     }
 
-    fn drawBackround(self: *Self, comptime n: u2) void {
+    fn drawBackground(self: *Self, comptime n: u2) void {
         // A Tile in a charblock is a byte, while a Screen Entry is a halfword
 
         const char_base = 0x4000 * @as(u32, self.bg[n].cnt.char_base.read());
@@ -325,10 +329,11 @@ pub const Ppu = struct {
 
         var i: u32 = 0;
         while (i < width) : (i += 1) {
-            if (!shouldDrawBackground(n, self.bldcnt, &self.scanline, i)) continue;
-
             const x = hofs + i;
 
+            const win_bounds = self.windowBounds(@truncate(u9, x), @truncate(u8, y));
+            if (!shouldDrawBackground(self, n, win_bounds, i)) continue;
+
             // Grab the Screen Entry from VRAM
             const entry_addr = screen_base + tilemapOffset(size, x, y);
             const entry = @bitCast(ScreenEntry, self.vram.read(u16, entry_addr));
@@ -353,7 +358,7 @@ pub const Ppu = struct {
 
             if (pal_id != 0) {
                 const bgr555 = self.palette.read(u16, pal_id * 2);
-                copyToBackgroundBuffer(n, self.bldcnt, &self.scanline, i, bgr555);
+                self.copyToBackgroundBuffer(n, win_bounds, i, bgr555);
             }
         }
     }
@@ -379,10 +384,10 @@ pub const Ppu = struct {
                 var layer: usize = 0;
                 while (layer < 4) : (layer += 1) {
                     self.drawSprites(@truncate(u2, layer));
-                    if (layer == self.bg[0].cnt.priority.read() and bg_enable & 1 == 1) self.drawBackround(0);
-                    if (layer == self.bg[1].cnt.priority.read() and bg_enable >> 1 & 1 == 1) self.drawBackround(1);
-                    if (layer == self.bg[2].cnt.priority.read() and bg_enable >> 2 & 1 == 1) self.drawBackround(2);
-                    if (layer == self.bg[3].cnt.priority.read() and bg_enable >> 3 & 1 == 1) self.drawBackround(3);
+                    if (layer == self.bg[0].cnt.priority.read() and bg_enable & 1 == 1) self.drawBackground(0);
+                    if (layer == self.bg[1].cnt.priority.read() and bg_enable >> 1 & 1 == 1) self.drawBackground(1);
+                    if (layer == self.bg[2].cnt.priority.read() and bg_enable >> 2 & 1 == 1) self.drawBackground(2);
+                    if (layer == self.bg[3].cnt.priority.read() and bg_enable >> 3 & 1 == 1) self.drawBackground(3);
                 }
 
                 // Copy Drawn Scanline to Frame Buffer
@@ -407,8 +412,8 @@ pub const Ppu = struct {
                 var layer: usize = 0;
                 while (layer < 4) : (layer += 1) {
                     self.drawSprites(@truncate(u2, layer));
-                    if (layer == self.bg[0].cnt.priority.read() and bg_enable & 1 == 1) self.drawBackround(0);
-                    if (layer == self.bg[1].cnt.priority.read() and bg_enable >> 1 & 1 == 1) self.drawBackround(1);
+                    if (layer == self.bg[0].cnt.priority.read() and bg_enable & 1 == 1) self.drawBackground(0);
+                    if (layer == self.bg[1].cnt.priority.read() and bg_enable >> 1 & 1 == 1) self.drawBackground(1);
                     if (layer == self.bg[2].cnt.priority.read() and bg_enable >> 2 & 1 == 1) self.drawAffineBackground(2);
                 }
 
@@ -486,7 +491,7 @@ pub const Ppu = struct {
                 while (i < width) : (i += 1) {
                     // If we're outside of the bounds of mode 5, draw the background colour
                     const bgr555 =
-                        if (scanline < m5_height and i < m5_width) self.vram.read(u16, vram_base + i * @sizeOf(u16)) else self.palette.getBackdrop();
+                        if (scanline < m5_height and i < m5_width) self.vram.read(u16, vram_base + i * @sizeOf(u16)) else self.palette.backdrop();
 
                     std.mem.writeIntNative(u32, self.framebuf.get(.Emulator)[fb_base + i * @sizeOf(u32) ..][0..@sizeOf(u32)], COLOUR_LUT[bgr555 & 0x7FFF]);
                 }
@@ -530,7 +535,94 @@ pub const Ppu = struct {
         }
 
         if (maybe_top) |top| return top;
-        return self.palette.getBackdrop();
+        return self.palette.backdrop();
+    }
+
+    fn copyToBackgroundBuffer(self: *Self, comptime n: u2, bounds: ?WindowBounds, i: usize, bgr555: u16) void {
+        if (self.bldcnt.mode.read() != 0b00) {
+            // Standard Alpha Blending
+            const a_layers = self.bldcnt.layer_a.read();
+            const is_blend_enabled = (a_layers >> n) & 1 == 1;
+
+            // If Alpha Blending is enabled and we've found an eligible layer for
+            // Pixel A, store the pixel in the bottom pixel buffer
+
+            const win_part = if (bounds) |win| blk: {
+                // Window Enabled
+                break :blk switch (win) {
+                    .win0 => self.win.in.w0_bld.read(),
+                    .win1 => self.win.in.w1_bld.read(),
+                    .out => self.win.out.out_bld.read(),
+                };
+            } else true;
+
+            if (win_part and is_blend_enabled) {
+                self.scanline.btm()[i] = bgr555;
+                return;
+            }
+        }
+
+        self.scanline.top()[i] = bgr555;
+    }
+
+    const WindowBounds = enum { win0, win1, out };
+
+    fn windowBounds(self: *Self, x: u9, y: u8) ?WindowBounds {
+        const win0 = self.dispcnt.win_enable.read() & 1 == 1;
+        const win1 = (self.dispcnt.win_enable.read() >> 1) & 1 == 1;
+        const winObj = self.dispcnt.obj_win_enable.read();
+
+        if (!(win0 or win1 or winObj)) return null;
+
+        if (win0 and self.win.inRange(0, x, y)) return .win0;
+        if (win1 and self.win.inRange(1, x, y)) return .win1;
+
+        return .out;
+    }
+
+    fn shouldDrawBackground(self: *Self, comptime n: u2, bounds: ?WindowBounds, i: usize) bool {
+        // If a pixel has been drawn on the top layer, it's because:
+        // 1. The pixel is to be blended with a pixel on the bottom layer
+        // 2. The pixel is not to be blended at all
+        // Also, if we find a pixel on the top layer we don't need to bother with this I think?
+        if (self.scanline.top()[i] != null) return false;
+
+        if (bounds) |win| {
+            switch (win) {
+                .win0 => if ((self.win.in.w0_bg.read() >> n) & 1 == 0) return false,
+                .win1 => if ((self.win.in.w1_bg.read() >> n) & 1 == 0) return false,
+                .out => if ((self.win.out.out_bg.read() >> n) & 1 == 0) return false,
+            }
+        }
+
+        if (self.scanline.btm()[i] != null) {
+            // The pixel found in the bottom layer is:
+            // 1. From a higher priority background
+            // 2. From a background that is marked for blending (Pixel A)
+
+            // If Alpha Blending isn't enabled, then we've already found a higher prio
+            // pixel, we can return early
+            if (self.bldcnt.mode.read() != 0b01) return false;
+
+            const b_layers = self.bldcnt.layer_b.read();
+
+            const win_part = if (bounds) |win| blk: {
+                // Window Enabled
+                break :blk switch (win) {
+                    .win0 => self.win.in.w0_bld.read(),
+                    .win1 => self.win.in.w1_bld.read(),
+                    .out => self.win.out.out_bld.read(),
+                };
+            } else true;
+
+            // If the Background is not marked for blending, we've already found
+            // a higher priority pixel, move on.
+
+            const is_blend_enabled = win_part and ((b_layers >> n) & 1 == 1);
+            if (!is_blend_enabled) return false;
+        }
+
+        return true;
     }
 
     // TODO: Comment this + get a better understanding
@@ -572,7 +664,7 @@ pub const Ppu = struct {
         // See if HBlank DMA is present and not enabled
 
         if (!self.dispstat.vblank.read())
-            pollBlankingDma(cpu.bus, .HBlank);
+            dma.onBlanking(cpu.bus, .HBlank);
 
         self.dispstat.hblank.set();
         self.sched.push(.HBlank, 68 * 4 -| late);
@@ -614,7 +706,7 @@ pub const Ppu = struct {
                 self.aff_bg[1].latchRefPoints();
 
                 // See if Vblank DMA is present and not enabled
-                pollBlankingDma(cpu.bus, .VBlank);
+                dma.onBlanking(cpu.bus, .VBlank);
             }
 
             if (scanline == 227) self.dispstat.vblank.unset();
@@ -623,158 +715,6 @@ pub const Ppu = struct {
     }
 };
 
-const Palette = struct {
-    const palram_size = 0x400;
-    const Self = @This();
-
-    buf: []u8,
-    allocator: Allocator,
-
-    fn init(allocator: Allocator) !Self {
-        const buf = try allocator.alloc(u8, palram_size);
-        std.mem.set(u8, buf, 0);
-
-        return Self{
-            .buf = buf,
-            .allocator = allocator,
-        };
-    }
-
-    fn deinit(self: *Self) void {
-        self.allocator.free(self.buf);
-        self.* = undefined;
-    }
-
-    pub fn read(self: *const Self, comptime T: type, address: usize) T {
-        const addr = address & 0x3FF;
-
-        return switch (T) {
-            u32, u16, u8 => std.mem.readIntSliceLittle(T, self.buf[addr..][0..@sizeOf(T)]),
-            else => @compileError("PALRAM: Unsupported read width"),
-        };
-    }
-
-    pub fn write(self: *Self, comptime T: type, address: usize, value: T) void {
-        const addr = address & 0x3FF;
-
-        switch (T) {
-            u32, u16 => std.mem.writeIntSliceLittle(T, self.buf[addr..][0..@sizeOf(T)], value),
-            u8 => {
-                const align_addr = addr & ~@as(u32, 1); // Aligned to Halfword boundary
-                std.mem.writeIntSliceLittle(u16, self.buf[align_addr..][0..@sizeOf(u16)], @as(u16, value) * 0x101);
-            },
-            else => @compileError("PALRAM: Unsupported write width"),
-        }
-    }
-
-    fn getBackdrop(self: *const Self) u16 {
-        return self.read(u16, 0);
-    }
-};
-
-const Vram = struct {
-    const vram_size = 0x18000;
-    const Self = @This();
-
-    buf: []u8,
-    allocator: Allocator,
-
-    fn init(allocator: Allocator) !Self {
-        const buf = try allocator.alloc(u8, vram_size);
-        std.mem.set(u8, buf, 0);
-
-        return Self{
-            .buf = buf,
-            .allocator = allocator,
-        };
-    }
-
-    fn deinit(self: *Self) void {
-        self.allocator.free(self.buf);
-        self.* = undefined;
-    }
-
-    pub fn read(self: *const Self, comptime T: type, address: usize) T {
-        const addr = Self.mirror(address);
-
-        return switch (T) {
-            u32, u16, u8 => std.mem.readIntSliceLittle(T, self.buf[addr..][0..@sizeOf(T)]),
-            else => @compileError("VRAM: Unsupported read width"),
-        };
-    }
-
-    pub fn write(self: *Self, comptime T: type, dispcnt: io.DisplayControl, address: usize, value: T) void {
-        const mode: u3 = dispcnt.bg_mode.read();
-        const idx = Self.mirror(address);
-
-        switch (T) {
-            u32, u16 => std.mem.writeIntSliceLittle(T, self.buf[idx..][0..@sizeOf(T)], value),
-            u8 => {
-                // Ignore write if it falls within the boundaries of OBJ VRAM
-                switch (mode) {
-                    0, 1, 2 => if (0x0001_0000 <= idx) return,
-                    else => if (0x0001_4000 <= idx) return,
-                }
-
-                const align_idx = idx & ~@as(u32, 1); // Aligned to a halfword boundary
-                std.mem.writeIntSliceLittle(u16, self.buf[align_idx..][0..@sizeOf(u16)], @as(u16, value) * 0x101);
-            },
-            else => @compileError("VRAM: Unsupported write width"),
-        }
-    }
-
-    fn mirror(address: usize) usize {
-        // Mirrored in steps of 128K (64K + 32K + 32K) (abcc)
-        const addr = address & 0x1FFFF;
-
-        // If the address is within 96K we don't do anything,
-        // otherwise we want to mirror the last 32K (addresses between 64K and 96K)
-        return if (addr < vram_size) addr else 0x10000 + (addr & 0x7FFF);
-    }
-};
-
-const Oam = struct {
-    const oam_size = 0x400;
-    const Self = @This();
-
-    buf: []u8,
-    allocator: Allocator,
-
-    fn init(allocator: Allocator) !Self {
-        const buf = try allocator.alloc(u8, oam_size);
-        std.mem.set(u8, buf, 0);
-
-        return Self{
-            .buf = buf,
-            .allocator = allocator,
-        };
-    }
-
-    fn deinit(self: *Self) void {
-        self.allocator.free(self.buf);
-        self.* = undefined;
-    }
-
-    pub fn read(self: *const Self, comptime T: type, address: usize) T {
-        const addr = address & 0x3FF;
-
-        return switch (T) {
-            u32, u16, u8 => std.mem.readIntSliceLittle(T, self.buf[addr..][0..@sizeOf(T)]),
-            else => @compileError("OAM: Unsupported read width"),
-        };
-    }
-
-    pub fn write(self: *Self, comptime T: type, address: usize, value: T) void {
-        const addr = address & 0x3FF;
-
-        switch (T) {
-            u32, u16 => std.mem.writeIntSliceLittle(T, self.buf[addr..][0..@sizeOf(T)], value),
-            u8 => return, // 8-bit writes are explicitly ignored
-            else => @compileError("OAM: Unsupported write width"),
-        }
-    }
-};
-
 const Window = struct {
     const Self = @This();
 
@@ -794,6 +734,25 @@ const Window = struct {
         };
     }
 
+    fn inRange(self: *const Self, comptime id: u1, x: u9, y: u8) bool {
+        const h = self.h[id];
+        const v = self.v[id];
+
+        const y1 = v.y1.read();
+        const y2 = if (y1 > v.y2.read()) 160 else std.math.min(160, v.y2.read());
+
+        if (y1 <= y and y < y2) {
+            // Within Y bounds
+            const x1 = h.x1.read();
+            const x2 = if (x1 > h.x2.read()) 240 else std.math.min(240, h.x2.read());
+
+            // Within X Bounds
+            return x1 <= x and x < x2;
+        }
+
+        return false;
+    }
+
     pub fn setH(self: *Self, value: u32) void {
         self.h[0].raw = @truncate(u16, value);
         self.h[1].raw = @truncate(u16, value >> 16);
@@ -1135,37 +1094,6 @@ fn alphaBlend(top: u16, btm: u16, bldalpha: io.BldAlpha) u16 {
     return (bld_b << 10) | (bld_g << 5) | bld_r;
 }
 
-fn shouldDrawBackground(comptime n: u2, bldcnt: io.BldCnt, scanline: *Scanline, i: usize) bool {
-    // If a pixel has been drawn on the top layer, it's because
-    // Either the pixel is to be blended with a pixel on the bottom layer
-    // or the pixel is not to be blended at all
-    // Consequentially, if we find a pixel on the top layer, there's no need
-    // to render anything I think?
-    if (scanline.top()[i] != null) return false;
-
-    if (scanline.btm()[i] != null) {
-        // The Pixel found in the Bottom layer is
-        // 1. From a higher priority
-        // 2. From a Backround that is marked for Blending (Pixel A)
-        //
-        // We now have to confirm whether this current Background can be used
-        // as Pixel B or not.
-
-        // If Alpha Blending isn't enabled, we've aready found a higher
-        // priority pixel to render. Move on
-        if (bldcnt.mode.read() != 0b01) return false;
-
-        const b_layers = bldcnt.layer_b.read();
-        const is_blend_enabled = (b_layers >> n) & 1 == 1;
-
-        // If the Background is not marked for blending, we've already found
-        // a higher priority pixel, move on.
-        if (!is_blend_enabled) return false;
-    }
-
-    return true;
-}
-
 fn shouldDrawSprite(bldcnt: io.BldCnt, scanline: *Scanline, x: u9) bool {
     if (scanline.top()[x] != null) return false;
 
@@ -1180,23 +1108,6 @@ fn shouldDrawSprite(bldcnt: io.BldCnt, scanline: *Scanline, x: u9) bool {
     return true;
 }
 
-fn copyToBackgroundBuffer(comptime n: u2, bldcnt: io.BldCnt, scanline: *Scanline, i: usize, bgr555: u16) void {
-    if (bldcnt.mode.read() != 0b00) {
-        // Standard Alpha Blending
-        const a_layers = bldcnt.layer_a.read();
-        const is_blend_enabled = (a_layers >> n) & 1 == 1;
-
-        // If Alpha Blending is enabled and we've found an eligible layer for
-        // Pixel A, store the pixel in the bottom pixel buffer
-        if (is_blend_enabled) {
-            scanline.btm()[i] = bgr555;
-            return;
-        }
-    }
-
-    scanline.top()[i] = bgr555;
-}
-
 fn copyToSpriteBuffer(bldcnt: io.BldCnt, scanline: *Scanline, x: u9, bgr555: u16) void {
     if (bldcnt.mode.read() != 0b00) {
         // Alpha Blending
@@ -1249,48 +1160,3 @@ const Scanline = struct {
         return self.layers[1];
     }
 };
-
-// Double Buffering Implementation
-const FrameBuffer = struct {
-    const Self = @This();
-
-    layers: [2][]u8,
-    buf: []u8,
-    current: u1,
-
-    allocator: Allocator,
-
-    // TODO: Rename
-    const Device = enum {
-        Emulator,
-        Renderer,
-    };
-
-    pub fn init(allocator: Allocator) !Self {
-        const framebuf_len = framebuf_pitch * height;
-        const buf = try allocator.alloc(u8, framebuf_len * 2);
-        std.mem.set(u8, buf, 0);
-
-        return .{
-            // Front and Back Framebuffers
-            .layers = [_][]u8{ buf[0..][0..framebuf_len], buf[framebuf_len..][0..framebuf_len] },
-            .buf = buf,
-            .current = 0,
-
-            .allocator = allocator,
-        };
-    }
-
-    fn deinit(self: *Self) void {
-        self.allocator.free(self.buf);
-        self.* = undefined;
-    }
-
-    pub fn swap(self: *Self) void {
-        self.current = ~self.current;
-    }
-
-    pub fn get(self: *Self, comptime dev: Device) []u8 {
-        return self.layers[if (dev == .Emulator) self.current else ~self.current];
-    }
-};

src/core/ppu/Oam.zig

@@ -0,0 +1,40 @@
+const std = @import("std");
+
+const Allocator = std.mem.Allocator;
+
+const buf_len = 0x400;
+const Self = @This();
+
+buf: []u8,
+allocator: Allocator,
+
+pub fn read(self: *const Self, comptime T: type, address: usize) T {
+    const addr = address & 0x3FF;
+
+    return switch (T) {
+        u32, u16, u8 => std.mem.readIntSliceLittle(T, self.buf[addr..][0..@sizeOf(T)]),
+        else => @compileError("OAM: Unsupported read width"),
+    };
+}
+
+pub fn write(self: *Self, comptime T: type, address: usize, value: T) void {
+    const addr = address & 0x3FF;
+
+    switch (T) {
+        u32, u16 => std.mem.writeIntSliceLittle(T, self.buf[addr..][0..@sizeOf(T)], value),
+        u8 => return, // 8-bit writes are explicitly ignored
+        else => @compileError("OAM: Unsupported write width"),
+    }
+}
+
+pub fn init(allocator: Allocator) !Self {
+    const buf = try allocator.alloc(u8, buf_len);
+    std.mem.set(u8, buf, 0);
+
+    return Self{ .buf = buf, .allocator = allocator };
+}
+
+pub fn deinit(self: *Self) void {
+    self.allocator.free(self.buf);
+    self.* = undefined;
+}

src/core/ppu/Palette.zig

@@ -0,0 +1,47 @@
+const std = @import("std");
+
+const Allocator = std.mem.Allocator;
+
+const buf_len = 0x400;
+const Self = @This();
+
+buf: []u8,
+allocator: Allocator,
+
+pub fn read(self: *const Self, comptime T: type, address: usize) T {
+    const addr = address & 0x3FF;
+
+    return switch (T) {
+        u32, u16, u8 => std.mem.readIntSliceLittle(T, self.buf[addr..][0..@sizeOf(T)]),
+        else => @compileError("PALRAM: Unsupported read width"),
+    };
+}
+
+pub fn write(self: *Self, comptime T: type, address: usize, value: T) void {
+    const addr = address & 0x3FF;
+
+    switch (T) {
+        u32, u16 => std.mem.writeIntSliceLittle(T, self.buf[addr..][0..@sizeOf(T)], value),
+        u8 => {
+            const align_addr = addr & ~@as(u32, 1); // Aligned to Halfword boundary
+            std.mem.writeIntSliceLittle(u16, self.buf[align_addr..][0..@sizeOf(u16)], @as(u16, value) * 0x101);
+        },
+        else => @compileError("PALRAM: Unsupported write width"),
+    }
+}
+
+pub fn init(allocator: Allocator) !Self {
+    const buf = try allocator.alloc(u8, buf_len);
+    std.mem.set(u8, buf, 0);
+
+    return Self{ .buf = buf, .allocator = allocator };
+}
+
+pub fn deinit(self: *Self) void {
+    self.allocator.free(self.buf);
+    self.* = undefined;
+}
+
+pub fn backdrop(self: *const Self) u16 {
+    return self.read(u16, 0);
+}

src/core/ppu/Vram.zig

@@ -0,0 +1,60 @@
+const std = @import("std");
+const io = @import("../bus/io.zig");
+
+const Allocator = std.mem.Allocator;
+
+const buf_len = 0x18000;
+const Self = @This();
+
+buf: []u8,
+allocator: Allocator,
+
+pub fn read(self: *const Self, comptime T: type, address: usize) T {
+    const addr = Self.mirror(address);
+
+    return switch (T) {
+        u32, u16, u8 => std.mem.readIntSliceLittle(T, self.buf[addr..][0..@sizeOf(T)]),
+        else => @compileError("VRAM: Unsupported read width"),
+    };
+}
+
+pub fn write(self: *Self, comptime T: type, dispcnt: io.DisplayControl, address: usize, value: T) void {
+    const mode: u3 = dispcnt.bg_mode.read();
+    const idx = Self.mirror(address);
+
+    switch (T) {
+        u32, u16 => std.mem.writeIntSliceLittle(T, self.buf[idx..][0..@sizeOf(T)], value),
+        u8 => {
+            // Ignore write if it falls within the boundaries of OBJ VRAM
+            switch (mode) {
+                0, 1, 2 => if (0x0001_0000 <= idx) return,
+                else => if (0x0001_4000 <= idx) return,
+            }
+
+            const align_idx = idx & ~@as(u32, 1); // Aligned to a halfword boundary
+            std.mem.writeIntSliceLittle(u16, self.buf[align_idx..][0..@sizeOf(u16)], @as(u16, value) * 0x101);
+        },
+        else => @compileError("VRAM: Unsupported write width"),
+    }
+}
+
+pub fn init(allocator: Allocator) !Self {
+    const buf = try allocator.alloc(u8, buf_len);
+    std.mem.set(u8, buf, 0);
+
+    return Self{ .buf = buf, .allocator = allocator };
+}
+
+pub fn deinit(self: *Self) void {
+    self.allocator.free(self.buf);
+    self.* = undefined;
+}
+
+fn mirror(address: usize) usize {
+    // Mirrored in steps of 128K (64K + 32K + 32K) (abcc)
+    const addr = address & 0x1FFFF;
+
+    // If the address is within 96K we don't do anything,
+    // otherwise we want to mirror the last 32K (addresses between 64K and 96K)
+    return if (addr < buf_len) addr else 0x10000 + (addr & 0x7FFF);
+}

src/main.zig

@@ -8,7 +8,7 @@ const Gui = @import("platform.zig").Gui;
 const Bus = @import("core/Bus.zig");
 const Arm7tdmi = @import("core/cpu.zig").Arm7tdmi;
 const Scheduler = @import("core/scheduler.zig").Scheduler;
-const FilePaths = @import("core/util.zig").FilePaths;
+const FilePaths = @import("util.zig").FilePaths;
 
 const Allocator = std.mem.Allocator;
 const log = std.log.scoped(.Cli);

src/platform.zig

@@ -5,9 +5,9 @@ const emu = @import("core/emu.zig");
 const Apu = @import("core/apu.zig").Apu;
 const Arm7tdmi = @import("core/cpu.zig").Arm7tdmi;
 const Scheduler = @import("core/scheduler.zig").Scheduler;
-const FpsTracker = @import("core/util.zig").FpsTracker;
+const FpsTracker = @import("util.zig").FpsTracker;
 
-const span = @import("core/util.zig").span;
+const span = @import("util.zig").span;
 
 const pitch = @import("core/ppu.zig").framebuf_pitch;
 const scale = @import("core/emu.zig").win_scale;

src/util.zig

@@ -1,9 +1,11 @@
 const std = @import("std");
 const builtin = @import("builtin");
 const Log2Int = std.math.Log2Int;
-const Arm7tdmi = @import("cpu.zig").Arm7tdmi;
+const Arm7tdmi = @import("core/cpu.zig").Arm7tdmi;
 
-const allow_unhandled_io = @import("emu.zig").allow_unhandled_io;
+const Allocator = std.mem.Allocator;
+
+const allow_unhandled_io = @import("core/emu.zig").allow_unhandled_io;
 
 // Sign-Extend value of type `T` to type `U`
 pub fn sext(comptime T: type, comptime U: type, value: T) T {
@@ -172,6 +174,7 @@ pub fn writeUndefined(log: anytype, comptime format: []const u8, args: anytype)
 
 pub const Logger = struct {
     const Self = @This();
+    const FmtArgTuple = std.meta.Tuple(&.{ u32, u32, u32, u32, u32, u32, u32, u32, u32, u32, u32, u32, u32, u32, u32, u32, u32, u32 });
 
     buf: std.io.BufferedWriter(4096 << 2, std.fs.File.Writer),
 
@@ -229,4 +232,43 @@ pub const Logger = struct {
     }
 };
 
-const FmtArgTuple = std.meta.Tuple(&.{ u32, u32, u32, u32, u32, u32, u32, u32, u32, u32, u32, u32, u32, u32, u32, u32, u32, u32 });
+// Double Buffering Implementation
+pub const FrameBuffer = struct {
+    const Self = @This();
+
+    layers: [2][]u8,
+    buf: []u8,
+    current: u1,
+
+    allocator: Allocator,
+
+    // TODO: Rename
+    const Device = enum { Emulator, Renderer };
+
+    pub fn init(allocator: Allocator, comptime len: comptime_int) !Self {
+        const buf = try allocator.alloc(u8, len * 2);
+        std.mem.set(u8, buf, 0);
+
+        return .{
+            // Front and Back Framebuffers
+            .layers = [_][]u8{ buf[0..][0..len], buf[len..][0..len] },
+            .buf = buf,
+            .current = 0,
+
+            .allocator = allocator,
+        };
+    }
+
+    pub fn deinit(self: *Self) void {
+        self.allocator.free(self.buf);
+        self.* = undefined;
+    }
+
+    pub fn swap(self: *Self) void {
+        self.current = ~self.current;
+    }
+
+    pub fn get(self: *Self, comptime dev: Device) []u8 {
+        return self.layers[if (dev == .Emulator) self.current else ~self.current];
+    }
+};