diff --git a/src/arm/cpu/arm/psr_transfer.zig b/src/arm/cpu/arm/psr_transfer.zig
index 0f1ae47..3280987 100644
--- a/src/arm/cpu/arm/psr_transfer.zig
+++ b/src/arm/cpu/arm/psr_transfer.zig
@@ -108,26 +108,21 @@ pub fn control(comptime InstrFn: type, comptime I: bool, comptime op: u6) InstrF
 
                     const rm = opcode & 0xF;
                     const rs = opcode >> 8 & 0xF;
-                    const rd_lo = opcode >> 12 & 0xF;
-                    const rd_hi = opcode >> 16 & 0xF;
+                    const rdlo = opcode >> 12 & 0xF;
+                    const rdhi = opcode >> 16 & 0xF;
 
-                    const left: i32 = @as(i16, @bitCast(@as(u16, @truncate(cpu.r[rm] >> 16 * X))));
-                    const right: i32 = @as(i16, @bitCast(@as(u16, @truncate(cpu.r[rs] >> 16 * Y))));
-
-                    // TODO: de-clutter this lmao
-                    const rdlo_val: i32 = @bitCast(cpu.r[rd_lo]);
+                    const left: i64 = @as(i16, @bitCast(@as(u16, @truncate(cpu.r[rm] >> 16 * X))));
+                    const right: i64 = @as(i16, @bitCast(@as(u16, @truncate(cpu.r[rs] >> 16 * Y))));
                     const product = left * right;
 
-                    // WHY DOESN'T THIS WORK????????
+                    const rdhi_val: i32 = @bitCast(cpu.r[rdhi]);
+                    const rdlo_val: i32 = @bitCast(cpu.r[rdlo]);
 
-                    cpu.r[rd_lo] = @bitCast(rdlo_val + product);
-                    cpu.r[rd_hi] = blk: {
-                        // FIXME: I think this has to do with correcting sign values?
-                        const offset_thing: i32 = @bitCast(if (product < 0) 0xFFFF_FFFF else @as(u32, 0));
+                    const accumulate = @as(i64, rdhi_val) << 32 | rdlo_val;
+                    const sum = product +% accumulate;
 
-                        const rdhi_val: i32 = @bitCast(cpu.r[rd_hi]);
-                        break :blk @bitCast(rdhi_val + offset_thing + @addWithOverflow(rdlo_val, product)[1]);
-                    };
+                    cpu.r[rdhi] = @bitCast(@as(i32, @truncate(sum >> 32)));
+                    cpu.r[rdlo] = @bitCast(@as(i32, @truncate(sum)));
                 },
                 0b01_1000, 0b01_1100 => { // SMLAW<y>
                     const Y = op >> 2 & 1;
@@ -178,16 +173,6 @@ pub fn control(comptime InstrFn: type, comptime I: bool, comptime op: u6) InstrF
             }
         }
 
-        // TODO: make generic on any integer?
-        inline fn carryFrom(left: i32, right: i32) u1 {
-            const _left: u32 = @bitCast(left);
-            const _right: u32 = @bitCast(right);
-
-            const sum = @as(u64, _left) + @as(u64, _right);
-
-            return @intCast(sum >> 32 & 1);
-        }
-
         inline fn msr(comptime R: bool, comptime imm: bool, cpu: *Arm32, opcode: u32) void {
             const field_mask: u4 = @truncate(opcode >> 16 & 0xF);
             const rm_idx = opcode & 0xF;