/*****************************************************************************

  The following code is derived, directly or indirectly, from the SystemC
  source code Copyright (c) 1996-2002 by all Contributors.
  All Rights reserved.

  The contents of this file are subject to the restrictions and limitations
  set forth in the SystemC Open Source License Version 2.3 (the "License");
  You may not use this file except in compliance with such restrictions and
  limitations. You may obtain instructions on how to receive a copy of the
  License at http://www.systemc.org/. Software distributed by Contributors
  under the License is distributed on an "AS IS" basis, WITHOUT WARRANTY OF
  ANY KIND, either express or implied. See the License for the specific
  language governing rights and limitations under the License.

 *****************************************************************************/

/*****************************************************************************

  fft.cpp - This is the implementation file for the synchronous process "fft".

  Original Author: Rashmi Goswami, Synopsys, Inc.

 *****************************************************************************/

/*****************************************************************************

  MODIFICATION LOG - modifiers, enter your name, affiliation, date and
  changes you are making here.

      Name, Affiliation, Date:
  Description of Modification:

 *****************************************************************************/


/* This is the implementation file for the synchronous process "fft" */
#include "systemc.h"
#include "fft.h"


//Function for butterfly computation

 void func_butterfly
    ( const fx_db& w_real   /* snps width 16 */,
      const fx_db& w_imag   /* snps width 16 */, 
      const fx_db& real1_in /* snps width 16 */,
      const fx_db& imag1_in /* snps width 16 */,
      const fx_db& real2_in /* snps width 16 */,
      const fx_db& imag2_in /* snps width 16 */,
      fx_db& real1_out /* snps width 16 */,
      fx_db& imag1_out /* snps width 16 */,
      fx_db& real2_out /* snps width 16 */,
      fx_db& imag2_out /* snps width 16 */
    )
 {

   // Variable declarations
	 #ifndef DOUBLE_FFT
     sc_int<17> tmp_real1;
     sc_int<17> tmp_imag1;
     sc_int<17> tmp_real2;
     sc_int<17> tmp_imag2;
     sc_int<34> tmp_real3;
     sc_int<34> tmp_imag3;
     #else
     fx_db tmp_real1;
     fx_db tmp_imag1;
     fx_db tmp_real2;
     fx_db tmp_imag2;
     fx_db tmp_real3;
     fx_db tmp_imag3;
     #endif
    // Begin Computation

    tmp_real1 = real1_in + real2_in; 

    // <s,6,10> = <s,5,10> + <s,5,10>
    tmp_imag1 = imag1_in + imag2_in;

    // <s,6,10> = <s,5,10> - <s,5,10>
    tmp_real2 = real1_in - real2_in;

    // <s,6,10> = <s,5,10> - <s,5,10>
    tmp_imag2 = imag1_in - imag2_in;

    //   <s,13,20> = <s,6,10>*<s,5,10> - <s,6,10>*<s,5,10>
    tmp_real3 = tmp_real2*w_real - tmp_imag2*w_imag;

    //   <s,13,20> = <s,6,10>*<s,5,10> - <s,6,10>*<s,5,10>
    tmp_imag3 = tmp_real2*w_imag + tmp_imag2*w_real; 

    // assign the sign-bit(MSB)      
#ifndef DOUBLE_FFT
    real1_out[15] = tmp_real1[16];
    imag1_out[15] = tmp_imag1[16];

    // assign the rest of the bits
    real1_out.range(14,0) = tmp_real1.range(14,0);
    imag1_out.range(14,0) = tmp_imag1.range(14,0);

   // assign the sign-bit(MSB)      
    real2_out[15] = tmp_real3[33];
    imag2_out[15] = tmp_imag3[33];          

   // assign the rest of the bits
    real2_out.range(14,0) = tmp_real3.range(24,10);
    imag2_out.range(14,0) = tmp_imag3.range(24,10);
#else
	real1_out=tmp_real1;
	imag1_out=tmp_imag1;
	real2_out=tmp_real3;
    imag2_out=tmp_imag3;
#endif
 }; // end func_butterfly

void fft::entry()
{ 
 // Variable Declarations
  fx_db real[16];
  fx_db imag[16];
  fx_db tmp_real;
  fx_db tmp_imag;
  short index;
  sc_int<6> N;  
  sc_int<4> M;
  sc_int<6> len;
  fx_db W_real[7];
  fx_db W_imag[7];
  fx_db w_real;
  fx_db w_imag;
  fx_db w_rec_real;
  fx_db w_rec_imag;
#ifndef DOUBLE_FFT
  sc_int<32> w_temp1;
  sc_int<32> w_temp2;
  sc_int<32> w_temp3;
  sc_int<32> w_temp4;
  sc_int<33> w_temp5;
  sc_int<33> w_temp6;
#else
  fx_db w_temp1;
  fx_db w_temp2;
  fx_db w_temp3;
  fx_db w_temp4;
  fx_db w_temp5;
  fx_db w_temp6;
#endif
  fx_db real1_in;
  fx_db imag1_in;
  fx_db real2_in;
  fx_db imag2_in;
  fx_db real1_out;
  fx_db imag1_out;
  fx_db real2_out;
  fx_db imag2_out;
  sc_int<4> stage;
  short i;
  short j;
  short index2;
  short windex;
  short incr;

  while(true)
  { data_req.write(false);
    data_ready.write(false);
    index = 0; 
     
    wait();
    //Read in the Sample values
      cout << endl << "Reading in the samples..." << endl;
      while( index < 16 )
      {
       data_req.write(true);	   
       wait_until(data_valid.delayed() == true);
	   cout << endl << "data_req_write" << endl;
       tmp_real = in_real.read();
       tmp_imag = in_imag.read();
       real[index] = tmp_real;
       imag[index] = tmp_imag;
       index++;
       data_req.write(false);
       wait();
      }
      index = 0;

 
       // Initialize
       M = 4; N = 16; 
       len = N >> 1;

       cout << "Computing..." << endl;
       // Calculate the W-values recursively
       // <'s'/'u',m,n>: is used in comments to denote a fixed point representation
       // 's'- signed, 'u'- unsigned, m - no. of integer bits, n - no. of fractional bits

       //  theta = 8.0*atan(1.0)/N; theta = 22.5 degree

       //  w_real =  cos(theta) = 0.92 (000000.1110101110) <s,5,10>
           w_real =  942;

       //  w_imag = -sin(theta) = -0.38(111111.1001111010) <s,5,10>
           w_imag = -389;

       //  w_rec_real = 1(0000001.0000000000)
	   w_rec_real = 1024;

       //  w_rec_real = 0(000000.0000000000)	 
           w_rec_imag = 0;

        unsigned short w_index;

        w_index = 0;    
        while(w_index < 7) 
        {
	  // <s,11,20> = <s,5,10> * <s,5,10>
	   w_temp1 = w_rec_real*w_real;
	   w_temp2 = w_rec_imag*w_imag;

          // <s,11,20> = <s,5,10> * <s,5,10>
	   w_temp3 = w_rec_real*w_imag;
	   w_temp4 = w_rec_imag*w_real;	 

	  // <s,6,10> = <s,5,10> - <s,5,10>
           w_temp5 = w_temp1 - w_temp2;

	  // <s,6,10> = <s,5,10> + <s,5,10>
	   w_temp6 = w_temp3 + w_temp4;
#ifndef DOUBLE_FFT 
	  // assign the sign-bit(MSB)
           W_real[w_index][15] = w_temp5[32];
           W_imag[w_index][15] = w_temp6[32];

	  // assign the rest of the bits
           W_real[w_index].range(14,0) = w_temp5.range(24,10);
           W_imag[w_index].range(14,0) = w_temp6.range(24,10);

	  
#else
	       W_real[w_index] = w_temp5;
           W_imag[w_index] = w_temp6;
#endif
		  // update w_rec.. values for the next iteration
	   w_rec_real = W_real[w_index];
	   w_rec_imag = W_imag[w_index];
	   w_index++;

        }

      //////////////////////////////////////////////////////////////////////////
      ///  Computation - 1D Complex DFT In-Place DIF Computation Algorithm  ////
      //////////////////////////////////////////////////////////////////////////

       stage = 0;
       len = N;
       incr = 1;

       while (stage < M) 
       { 
	 len = len >> 1;
 
        //First Iteration :  Simple calculation, with no multiplies
          i = 0;
          while(i < N)
          {
             index =  i; index2 = i + len; 

             tmp_real = real[index] + real[index2];
             tmp_imag = imag[index] + imag[index2];

             real[index2] = (real[index] - real[index2]);
             imag[index2] = (imag[index] - imag[index2]);

             real[index] = tmp_real;
             imag[index] = tmp_imag;
	    
	     i = i + (len << 1);   
          }

        //Remaining Iterations: Use Stored W
         j = 1; windex = incr - 1;
        // This loop executes N/2 times at the first stage, N/2 times at the second.. once at last stage
         while (j < len)
         {
            i = j; 
            while (i < N)
            {
              index = i;
              index2 = i + len;

	      // Read in the data and twiddle factors
	      w_real  = W_real[windex];
	      w_imag  = W_imag[windex];

              real1_in = real[index];
	      imag1_in = imag[index];
              real2_in = real[index2];
	      imag2_in = imag[index2];

              // Call butterfly computation function	     
	      func_butterfly(w_real, w_imag, real1_in, imag1_in, real2_in, imag2_in, real1_out, imag1_out, real2_out, imag2_out);

	      // Store back the results
              real[index]  = real1_out;
              imag[index]  = imag1_out; 
              real[index2] = real2_out;
              imag[index2] = imag2_out; 

              i = i + (len << 1);
            }
            windex = windex + incr;
            j++;
         }
          stage++;
          incr = incr << 1;
       } 
           
     //////////////////////////////////////////////////////////////////////////   
     //Writing out the normalized transform values in bit reversed order
     //////////////////////////////////////////////////////////////////////////

      sc_uint<4> bits_i;
      sc_uint<4> bits_index;
      fx_db real1;
      fx_db imag1;
      bits_i = 0;
      bits_index = 0;
      i = 0;

      cout << "Writing the transform values..." << endl;
      while( i < 16)
      {
       bits_i = i;
       bits_index[3]= bits_i[0];
       bits_index[2]= bits_i[1];
       bits_index[1]= bits_i[2];
       bits_index[0]= bits_i[3];
       index = bits_index;
       real1 = real[index];
       imag1 = imag[index];
       out_real.write(real1); 
       out_imag.write(imag1); 
       data_ready.write(true);
       wait_until(data_ack.delayed() == true);
       data_ready.write(false);
       i++;
       wait();
      }
      index = 0; 
      cout << "Done..." << endl;
  }      
}// end entry() function