Fixed bug in adder/subtracter where code assuming width was the bit width was not updated when moving to the 2**N approach required by the assignment. Updated adder/subtracter testbench to correctly instantiate the asser/subtracter and removed unused variable

modules/asu.sv

@@ -7,11 +7,11 @@ module asu #(parameter width = 2)(
     always_comb begin
         // A + (sub) ? B : -B
         // Adds an extra bit to all operands to calculate the carry bit. Not optimal for hardware synthesis
-        operand = (B ^ {(width){op}});
-        {COUT, OUT_1} = A + operand + {{(width-1){1'b0}}, op};
+        operand = (B ^ {(2**width){op}});
+        {COUT, OUT_1} = A + operand + {{(2**width-1){1'b0}}, op};
     end
 
     // Need to do this outside of the always_comb block because parameter-based bitslicing/indexing isn't supported?
-    // Cin for width-2'th bit to later be xor'ed with COUT to determine if a signed overflow occured for the MIPS architecture
-    assign OUT_CIN = A[width-1] ^ operand[width-1] ^ OUT_1[width-1];
+    // Cin for final bit in adder to later be xor'ed with COUT to determine if a signed overflow occured for the MIPS architecture
+    assign OUT_CIN = A[2**width-1] ^ operand[2**width-1] ^ OUT_1[2**width-1];
 endmodule

tests/asu_tb.sv

@@ -10,10 +10,8 @@ module asu_tb #(width=8)();
     logic [width-1:0] C_exp;
     logic [0:0] Cout_exp;
     logic [0:0] Cin_exp;   // TODO
-    
-    logic [0:0] Signed_overflow;
 
-    asu #(width) dut(op, A, B, C, Cout, Cin);
+    asu #($clog2(width)) dut(op, A, B, C, Cout, Cin);
 
     integer error_count;
     initial begin