> When compiling my program
Which one?

Pls read your own post WITHOUT ANY knowledge about your problem. You 
will see: there is not nearly enough information for any hint!

And: do you know Google?
http://www.google.com/search?q=Library+lpm+not+found

says : ''cannot find expanded name "lpm.lpm_components".

I've looking on google and apparently i need a folder termed as 
"220model"?

Here is the code i want to simulate :

LIBRARY lpm;
USE lpm.lpm_components.ALL;

LIBRARY ieee;
USE ieee.std_logic_1164.ALL;
USE ieee.std_logic_arith.ALL;

PACKAGE mul_package is -- user defined components
  COMPONENT ccmul
    GENERIC (W2 : INTEGER := 17; --multiplier bit width
              W1 : INTEGER := 9; --bit width C+S SUM
              W   : INTEGER := 8); -- Input bit width

              PORT
              (clk : IN STD_LOGIC; -- clock for the output register
              x_in, y_in, c_in : IN STD_LOGIC_VECTOR(W-1 DOWNTO 0);
              --INPUTS
              cps_in, cms_in : IN STD_LOGIC_VECTOR(W-1 DOWNTO 0);
              --INPUTS
              r_out, i_out : OUT STD_LOGIC_VECTOR(W-1 DOWNTO 0));
              --RESULTS
            END COMPONENT;
          END mul_package;

LIBRARY work;
USE work.mul_package.ALL;

LIBRARY ieee;
USE ieee.std-logic_1164.ALL;
USE ieee.std_logic_arith.ALL;
LIBRARY lpm;
USE lpm.lpm_components.ALL;

LIBRARY ieee;
USE ieee.std_logic_1164.ALL;
USE ieee.std_logic_arith.ALL;
USE ieee.std_logic_unsigned.ALL;

ENTITY bfproc IS
  GENERIC ( W2 : INTEGER := 17;  --multiplier bit width
  W1 : INTEGER := 9;
  W : INTEGER := 8);
  PORT
  (clk    :STD_LOGIC;
  Are_in, Aim_in, c_in, --8 bit inputs
  Bre_in, Bim_in : IN STD_LOGIC_VECTOR(W-1 DOWNTO 0);
  cps_in, cms_in : IN STD_LOGIC_VECTOR(W1-1 DOWNTO 0); --9 bits 
coefficients

  Dre_out, Dim_out, --8bit results
  Ere_out , Eim_out : OUT STD_LOGIC_VECTOR(W1 DOWNTO 0));
END bfproc;

ARCHITECTURE flex OF bfproc IS
  SIGNAL dif_re, dif_im         --Bf out
  : STD_LOGIC_VECTOR(W-1 DOWNTO 0);
  SIGNAL Are, Aim, Bre, Bim : integer RANGE -128 TO 127;
  --inputs as integers
  SIGNAL c : STD_LOGIC_VECTOR(W-1 DOWNTO 0);
  --input
  SIGNAL cps, cms : STD_LOGIC_VECTOR(W1 DOWNTO 0);
  --RESULTS

BEGIN
  PROCESS --compute the additions of the butterfly using
    BEGIN -- integers and store inputs in flip flops
      WAIT UNTIL clk = '1';
      Are <= CONV_INTEGER (Are_in);
      Aim <= CONV_INTEGER(Aim_in);
      Bre <= CONV_INTEGER(Bre_in);
      Bim <= CONV_INTEGER(Bim_in);
      c <= c_in; --load from memory cos
      cps <= cps_in; --load from memory cos+sin
      cms <= cms_in; --Load from memory cos-sin
      Dre_out <= CONV_STD_LOGIC VECTOR (Are + Bre )/2, W);
      Dim_out <= CONV_STD_LOGIC_VECTOR(Aim + Bim)/2, W);
    END PROCESS;

    --No FF because butterfly difference ''diff'' is not an output port

    PROCESS (Are, Bre, Aim, Bim)
      BEGIN
        dif_re <= CONV_STD_LOGIC_VECTOR(Are/2 - Bre/2, 8);
        dif_im <= CONV_STD_LOGIC_VECTOR(Aim/2 - Bim/2,8);
      END PROCESS;

1

USE ieee.std-logic_1164.ALL;

Here is a typo, too!

> you have to compile the xilinx libaries
Xilinx? LPM are "Megafunctions" from Altera...     :-o
http://www.altera.com/support/software/eda_maxplus2/synopsys/fpgaex/fpxlpm.html