I want to design an 8-bit ALU. I have written this code, but when I simulate it, the output has x value. Why did it happen?
module eightBitAlu(clk, a, b,si,ci, opcode,outp);
input clk;
input [7:0] a, b;
input [2:0] opcode;
input si;
input ci;
output reg [7:0] outp;
always @(posedge clk)
begin
case (opcode)
3'b000: outp <= a - b;
3'b000 : outp <= a + b;
3'b001 : outp =0;
3'b010 : outp <= a & b;
3'b011 : outp <= a | b;
3'b100 : outp <= ~a;
endcase
end
endmodule
This is my test module:
module test_8bitAlu();
reg clk=0,a=3,b=1,si=0,ci=0,opcode=1;
eightBitAlu alu(clk, a, b,si,ci, opcode,outp);
initial begin
#200 clk=1;
#200 opcode=0;
#200 opcode=2;
#200 opcode=3;
#200 opcode=4;
#200;
end
endmodule
a and b are only 1 bit wide leaving the top 7 bits of your input ports un-driven.
reg clk=0,a=3,b=1,si=0,ci=0,opcode=1;
is equivalent to :
reg clk = 0;
reg a = 3;
reg b = 1;
reg si = 0;
reg ci = 0;
reg opcode = 1;
What you need is:
reg clk = 0;
reg [7:0] a = 3;
reg [7:0] b = 1;
reg si = 0;
reg ci = 0;
reg [2:0] opcode = 1;
wire [7:0] outp;
Further improvemnets would be to include the width on the integer assignment ie:
reg clk = 1'd0;
reg [7:0] a = 8'd3;
b for binary, d for decimal, o for octal and h for hexadecimal in width'formatValue
outp if not defined will be an implicit 1 bit wire.
Your clock in the testharness also only has 1 positive edge. You may prefer to define your clock as:
initial begin
clk = 1'b0;
forever begin
#100 clk = ~clk;
end
end
A complete version of the above is demonstrated at EDAplayground.