feat(cpu): Implement Multiply Long ARM instructions

src/bus/io.zig

@@ -21,7 +21,7 @@ pub const Io = struct {
             .vcount = .{ .raw = 0x0000 },
             .ime = false,
             .ie = .{ .raw = 0x0000 },
-            .keyinput = .{ .raw = 0x01FF },
+            .keyinput = .{ .raw = 0x03FF },
         };
     }
 

src/cpu.zig

@@ -18,6 +18,7 @@ const branch = @import("cpu/arm/branch.zig").branch;
 const branchAndExchange = @import("cpu/arm/branch.zig").branchAndExchange;
 const softwareInterrupt = @import("cpu/arm/software_interrupt.zig").softwareInterrupt;
 const multiply = @import("cpu/arm/multiply.zig").multiply;
+const multiplyLong = @import("cpu/arm/multiply_long.zig").multiplyLong;
 
 // THUMB Instruction Groups
 const format1 = @import("cpu/thumb/format1.zig").format1;
@@ -533,6 +534,14 @@ fn armPopulate() [0x1000]ArmInstrFn {
                 lut[i] = multiply(A, S);
             }
 
+            if (i >> 7 & 0x1F == 0b00001 and i & 0xF == 0b1001) {
+                const U = i >> 6 & 1 == 1;
+                const A = i >> 5 & 1 == 1;
+                const S = i >> 4 & 1 == 1;
+
+                lut[i] = multiplyLong(U, A, S);
+            }
+
             if (i >> 10 & 0x3 == 0b01) {
                 const I = i >> 9 & 1 == 1;
                 const P = i >> 8 & 1 == 1;

src/cpu/arm/multiply_long.zig

@@ -0,0 +1,38 @@
+const std = @import("std");
+
+const Bus = @import("../../Bus.zig");
+const Arm7tdmi = @import("../../cpu.zig").Arm7tdmi;
+const InstrFn = @import("../../cpu.zig").ArmInstrFn;
+
+pub fn multiplyLong(comptime U: bool, comptime A: bool, comptime S: bool) InstrFn {
+    return struct {
+        fn inner(cpu: *Arm7tdmi, _: *Bus, opcode: u32) void {
+            const rd_hi = opcode >> 16 & 0xF;
+            const rd_lo = opcode >> 12 & 0xF;
+            const rs = opcode >> 8 & 0xF;
+            const rm = opcode & 0xF;
+
+            if (U) {
+                // Signed (WHY IS IT U THEN?)
+                var result: i64 = @as(i64, @bitCast(i32, cpu.r[rm])) * @as(i64, @bitCast(i32, cpu.r[rs]));
+                if (A) result += @bitCast(i64, @as(u64, cpu.r[rd_hi]) << 32 | @as(u64, cpu.r[rd_lo]));
+
+                cpu.r[rd_hi] = @bitCast(u32, @truncate(i32, result >> 32));
+                cpu.r[rd_lo] = @bitCast(u32, @truncate(i32, result));
+            } else {
+                // Unsigned
+                var result: u64 = @as(u64, cpu.r[rm]) * @as(u64, cpu.r[rs]);
+                if (A) result += @as(u64, cpu.r[rd_hi]) << 32 | @as(u64, cpu.r[rd_lo]);
+
+                cpu.r[rd_hi] = @truncate(u32, result >> 32);
+                cpu.r[rd_lo] = @truncate(u32, result);
+            }
+
+            if (S) {
+                cpu.cpsr.z.write(cpu.r[rd_hi] == 0 and cpu.r[rd_lo] == 0);
+                cpu.cpsr.n.write(cpu.r[rd_hi] >> 31 & 1 == 1);
+                // C and V are set to meaningless values
+            }
+        }
+    }.inner;
+}