/* -*- c++ -*- */

/*

 * Copyright 2002 Free Software Foundation, Inc.

 * 

 * This file is part of GNU Radio

 * 

 * GNU Radio is free software; you can redistribute it and/or modify

 * it under the terms of the GNU General Public License as published by

 * the Free Software Foundation; either version 3, or (at your option)

 * any later version.

 * 

 * GNU Radio is distributed in the hope that it will be useful,

 * but WITHOUT ANY WARRANTY; without even the implied warranty of

 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the

 * GNU General Public License for more details.

 * 

 * You should have received a copy of the GNU General Public License

 * along with GNU Radio; see the file COPYING.  If not, write to

 * the Free Software Foundation, Inc., 51 Franklin Street,

 * Boston, MA 02110-1301, USA.

 */

/*

 * FIXME.  This code is virtually identical to qa_gr_fir_?CC.cc

 *   Kludge up some kind of macro to handle the minor differences.

 */

#ifdef HAVE_CONFIG_H

#include <config.h>

#endif

#include <gr_types.h>

typedef gr_complex        i_type;

typedef gr_complex        o_type;

typedef float                tap_type;

typedef        gr_complex        acc_type;

#include <qa_gr_fir_ccf.h>

#include <gr_fir_ccf.h>

#include <gr_fir_util.h>

#include <string.h>

#include <iostream>

#include <cmath>

#include <gr_types.h>

#include <cppunit/TestAssert.h>

#include <random.h>

#include <malloc16.h>

#include <string.h>

using std::vector;

#define        ERR_DELTA        (1e-5)

#define        NELEM(x) (sizeof (x) / sizeof (x[0]))

//

// typedef for something logically "pointer to constructor".

// there may be a better way, please let me know...

//

typedef gr_fir_ccf* (*fir_maker_t)(const std::vector<tap_type> &taps);

static float

uniform ()

{

  return 2.0 * ((float) random () / RANDOM_MAX - 0.5);        // uniformly (-1, 1)

}

static void

random_floats (float *buf, unsigned n)

{

  for (unsigned i = 0; i < n; i++)

    buf[i] = (float) rint (uniform () * 32767);

}

static void

random_complex (gr_complex *buf, unsigned n)

{

  for (unsigned i = 0; i < n; i++){

    float re = rint (uniform () * 32767);

    float im = rint (uniform () * 32767);

    buf[i] = gr_complex (re, im);

  }

}

static o_type

ref_dotprod (const i_type input[], const tap_type taps[], int ntaps)

{

  acc_type        sum = 0;

  for (int i = 0; i < ntaps; i++)

    sum += input[i] * taps[ntaps - i - 1];

  return sum;

}

//

// Test for ntaps in [0,9], and input lengths in [0,17].

// This ensures that we are building the shifted taps correctly,

// and exercises all corner cases on input alignment and length.

//

static void

test_random_io (fir_maker_t maker)  

{

  const int        MAX_TAPS        = 9;

  const int        OUTPUT_LEN        = 17;

  const int        INPUT_LEN        = MAX_TAPS + OUTPUT_LEN;

  // Our SIMD ccc kernel requires that the complex input be 64-bit (8-byte) aligned.

  //i_type         input[INPUT_LEN];

  i_type       *input = (i_type *)malloc16Align(INPUT_LEN * sizeof(i_type));

  o_type         expected_output[OUTPUT_LEN];

  o_type         actual_output[OUTPUT_LEN];

  tap_type        taps[MAX_TAPS];

  srandom (0);        // we want reproducibility

  for (int n = 0; n <= MAX_TAPS; n++){

    for (int ol = 0; ol <= OUTPUT_LEN; ol++){

      // cerr << "@@@ n:ol " << n << ":" << ol << endl;

      // build random test case

      random_complex (input, INPUT_LEN);

      random_floats (taps, MAX_TAPS);

      // compute expected output values

      for (int o = 0; o < ol; o++){

        expected_output[o] = ref_dotprod (&input[o], taps, n);

      }

      // build filter

      vector<tap_type> f1_taps (&taps[0], &taps[n]);

      gr_fir_ccf *f1 = maker (f1_taps);

      // zero the output, then do the filtering

      memset (actual_output, 0, sizeof (actual_output));

      f1->filterN (actual_output, input, ol);

      // check results

      //

      // we use a sloppy error margin because on the x86 architecture,

      // our reference implementation is using 80 bit floating point

      // arithmetic, while the SSE version is using 32 bit float point

      // arithmetic.

      for (int o = 0; o < ol; o++){

        CPPUNIT_ASSERT_COMPLEXES_EQUAL(expected_output[o], actual_output[o],

                                       abs (expected_output[o]) * ERR_DELTA);

      }

      delete f1;

    }

  }

  free16Align(input);

}

static void

for_each (void (*f)(fir_maker_t))

{

  std::vector<gr_fir_ccf_info>                info;

  gr_fir_util::get_gr_fir_ccf_info (&info);        // get all known ccf implementations 

  for (std::vector<gr_fir_ccf_info>::iterator p = info.begin ();

       p != info.end ();

       ++p){

    std::cerr << " [" << p->name << "]";

    f (p->create);

  }

  std::cerr << std::endl;

}

void

qa_gr_fir_ccf::t1 ()

{

  for_each (test_random_io);

}