multiplier

[Scofield]

Hello.
I need to write a multiplier in VHDL which multiplies two 32 bit registers,
assuming that the result is never bigger than 32 bits.

we used the following code:

ENTITY mult_rg IS
PORT (
d_in1: IN STD_LOGIC_VECTOR (31 DOWNTO 0);
d_in2: IN STD_LOGIC_VECTOR (31 DOWNTO 0);
d_out: OUT STD_LOGIC_VECTOR (31 DOWNTO 0)
);
END mult_rg;

ARCHITECTURE rtl OF mult_rg IS
SIGNAL a_int, b_int: SIGNED (31 downto 0);
SIGNAL pdt_int: SIGNED (31 downto 0);

BEGIN
a_int <= SIGNED (d_in1);
b_int <= SIGNED (d_in2);
pdt_int <= a_int * b_int;
d_out <= STD_LOGIC_VECTOR(pdt_int);
END rtl;
​

but I got the following error messege:

Length of expected is 32; length of actual is 64.

What should I do?
thanx.

 

[Jerrie85]

when you multiply two 32 bit numbers, the resultant is always 2 * 32 bits long at max 64 bits long ...

u can test this out by doing (2^32 - 1)(2^32-1) = (2^64 - 2^33 - 2)

i multiplied the two largest 32 bit numbers possible..and as u can see, the resultant is a 64 bit number

hence, change your d_out to (63 downto 0)...that'll fix it