/* $Id$ */
// Copyright (C) 2002, International Business Machines
// Corporation and others.  All Rights Reserved.
// This code is licensed under the terms of the Eclipse Public License (EPL).

#include "CoinPragma.hpp"

#include "AbcCommon.hpp"
#include "ClpSimplex.hpp"
#ifdef ABC_INHERIT
#include "AbcSimplex.hpp"
#endif
#ifndef ABC_INHERIT
CLPLIB_EXPORT
void ClpMain0(ClpSimplex *models);
CLPLIB_EXPORT
int ClpMain1(int argc, const char *argv[], ClpSimplex *model);
#else
CLPLIB_EXPORT
void ClpMain0(AbcSimplex *models);
CLPLIB_EXPORT
int ClpMain1(int argc, const char *argv[], AbcSimplex *model);
#endif
//#define CILK_TEST
#ifdef CILK_TEST
static void cilkTest();
#endif
//#define LAPACK_TEST
/*
  Somehow with some BLAS we get multithreaded by default
  For 99.99% of problems this is not a good idea.
  The openblas_set_num_threads(1) seems to work even with other blas
 */
#if CLP_USE_OPENBLAS
extern "C" {
void openblas_set_num_threads(int num_threads);
}
#endif
#ifdef LAPACK_TEST
#include "/include/lapacke.h"
#ifndef COIN_FACTORIZATION_DENSE_CODE
#define COIN_FACTORIZATION_DENSE_CODE 1
#endif
#if COIN_FACTORIZATION_DENSE_CODE == 1
// using simple lapack interface
extern "C" {
void openblas_set_num_threads(int num_threads);
#if 0
  /** LAPACK Fortran subroutine DGETRF. */
  void LAPACK_dgetrf(int * m, int *n,
                               double *A, int *ldA,
                               int * ipiv, int *info);
  /** LAPACK Fortran subroutine DGETRS. */
  void LAPACK_dgetrs(char *trans, int *n,
                               int *nrhs, const double *A, int *ldA,
		     int * ipiv, double *B, int *ldB, int *info);
  //	LAPACK_dgetrf(&N, &N, m, &LDA,ipiv, &info);
#endif
}
int test_lapack(int n)
{
  int *ipiv;
  int info;
  int i, j;
  double *m, *x, *y;

  int LDB, LDA, N, NRHS;
  char transp = 'N';

  m = (double *)malloc(sizeof(double) * n * n);
  x = (double *)malloc(sizeof(double) * n);
  y = (double *)malloc(sizeof(double) * n);
  ipiv = (int *)malloc(sizeof(int) * n);

  for (i = 0; i < n; ++i) {
    x[i] = 1.0;
    for (j = 0; j < n; ++j) {
      m[i * n + j] = (rand() % 100 + 1) / 10.0;
      //      printf("m[%d,%d]=%lf\n",i,j, m[i*n+j]);
    }
  }

  /* test cblas.h */
  //cblas_dgemv(CblasColMajor, CblasNoTrans, n, n, 1.0, m, n,
  //	      x, 1, 0.0, y, 1);

  //  for (i=0; i<n; ++i)  printf("x[%d]=%lf\n",i, x[i]);
  //for (i=0; i<n; ++i)  printf("y[%d]=%lf\n",i, y[i]);

  LDB = n;
  LDA = n;
  N = n;
  NRHS = 1;
  info = 0;

  LAPACK_dgetrf(&N, &N, m, &LDA, ipiv, &info);

  if (info != 0)
    fprintf(stderr, "dgetrf failure with error %d\n", info);

  LAPACK_dgetrs(&transp, &N, &NRHS, m, &LDA, ipiv, y, &LDB, &info);

  if (info != 0)
    fprintf(stderr, "failure with error %d\n", info);
  //  for (i=0; i<n; ++i) printf("%lf\n", y[i]);

  free(m);
  free(x);
  free(y);
  free(ipiv);
  return 0;
}
#elif COIN_FACTORIZATION_DENSE_CODE == 2
// C interface
enum CBLAS_ORDER { CblasRowMajor = 101,
  CblasColMajor = 102 };
enum CBLAS_TRANSPOSE { CblasNoTrans = 111,
  CblasTrans = 112 };
extern "C" {
int clapack_dgetrf(const enum CBLAS_ORDER Order, const int M, const int N, double *A, const int lda, int *ipiv);
int clapack_dgetrs(const enum CBLAS_ORDER Order,
  const enum CBLAS_TRANSPOSE Trans,
  const int N, const int NRHS,
  const double *A, const int lda, const int *ipiv, double *B,
  const int ldb);
}
int test_lapack(int n)
{
  int *ipiv;
  int info;
  int i, j;
  double *m, *x, *y;

  int LDB, LDA, N, NRHS;
  char transp = 'N';

  m = (double *)malloc(sizeof(double) * n * n);
  x = (double *)malloc(sizeof(double) * n);
  y = (double *)malloc(sizeof(double) * n);
  ipiv = (int *)malloc(sizeof(int) * n);

  for (i = 0; i < n; ++i) {
    x[i] = 1.0;
    for (j = 0; j < n; ++j) {
      m[i * n + j] = (rand() % 100 + 1) / 10.0;
      //      printf("m[%d,%d]=%lf\n",i,j, m[i*n+j]);
    }
  }

  /* test cblas.h */
  //cblas_dgemv(CblasColMajor, CblasNoTrans, n, n, 1.0, m, n,
  //	      x, 1, 0.0, y, 1);

  //  for (i=0; i<n; ++i)  printf("x[%d]=%lf\n",i, x[i]);
  //for (i=0; i<n; ++i)  printf("y[%d]=%lf\n",i, y[i]);

  LDB = n;
  LDA = n;
  N = n;
  NRHS = 1;
  info = clapack_dgetrf(CblasColMajor, n, n, m, n, ipiv);

  if (info != 0)
    fprintf(stderr, "dgetrf failure with error %d\n", info);

  clapack_dgetrs(CblasColMajor, CblasNoTrans, n, 1, m, n, ipiv, y, n);

  if (info != 0)
    fprintf(stderr, "failure with error %d\n", info);
  //  for (i=0; i<n; ++i) printf("%lf\n", y[i]);

  free(m);
  free(x);
  free(y);
  free(ipiv);
  return 0;
}
#endif
#endif
//#define CLP_MALLOC_STATISTICS
#ifdef CLP_MALLOC_STATISTICS
#include <malloc.h>
#include <exception>
#include <new>
static double malloc_times = 0.0;
static double malloc_total = 0.0;
static int malloc_amount[] = { 0, 32, 128, 256, 1024, 4096, 16384, 65536, 262144, COIN_INT_MAX };
static int malloc_n = 10;
double malloc_counts[10] = { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 };
bool malloc_counts_on = true;
void *operator new(size_t size) throw(std::bad_alloc)
{
  malloc_times++;
  malloc_total += size;
  int i;
  for (i = 0; i < malloc_n; i++) {
    if ((int)size <= malloc_amount[i]) {
      malloc_counts[i]++;
      break;
    }
  }
  if (size > 100000) {
    printf("allocating %ld bytes\n", size);
  }
  void *p = malloc(size);
  return p;
}
void operator delete(void *p) throw()
{
  free(p);
}
static void malloc_stats2()
{
  double average = malloc_total / malloc_times;
  printf("count %g bytes %g - average %g\n", malloc_times, malloc_total, average);
  for (int i = 0; i < malloc_n; i++)
    printf("%g ", malloc_counts[i]);
  printf("\n");
  malloc_times = 0.0;
  malloc_total = 0.0;
  memset(malloc_counts, 0, sizeof(malloc_counts));
  // print results
}
#endif //CLP_MALLOC_STATISTICS
int
#if defined(_MSC_VER)
  __cdecl
#endif // _MSC_VER
  main(int argc, const char *argv[])
{
#ifdef CILK_TEST
  cilkTest();
#endif
#if CLP_USE_OPENBLAS
  openblas_set_num_threads(CLP_USE_OPENBLAS);
#endif
#ifdef LAPACK_TEST
  //void openblas_set_num_threads(int num_threads);
  openblas_set_num_threads(1);
  if (argc < 2) {
    printf("Error - need size of matrix for lapack test\n");
    return 1;
  }
  int n = atoi(argv[1]);
  printf("n=%d\n", n);
  if (argc > 2) {
    int nThreads = atoi(argv[2]);
    printf("Using %d threads\n", nThreads);
    openblas_set_num_threads(nThreads);
  }
  test_lapack(n);
  return 0;
#endif
#ifndef ABC_INHERIT
  ClpSimplex *models = new ClpSimplex[1];
#else
  AbcSimplex *models = new AbcSimplex[1];
#endif
  std::cout << "Coin LP version " << CLP_VERSION
            << ", build " << __DATE__ << std::endl;
  // Print command line
  if (argc > 1) {
    printf("command line - ");
    for (int i = 0; i < argc; i++)
      printf("%s ", argv[i]);
    printf("\n");
  }
  ClpMain0(models);
  int returnCode = ClpMain1(argc, argv, models);
  delete[] models;
  return returnCode;
}
/*
  Version 1.00.00 October 13 2004.
  1.00.01 October 18.  Added basis handling helped/prodded by Thorsten Koch.
  Also modifications to make faster with sbb (I hope I haven't broken anything).
  1.00.02 March 21 2005.  Redid ClpNonLinearCost to save memory also redid
  createRim to try and improve cache characteristics.
  1.00.03 April 8 2005.  Added Volume algorithm as crash and made code more
  robust on testing.  Also added "either" and "tune" algorithm.
  1.01.01 April 12 2005.  Decided to go to different numbering.  Backups will
  be last 2 digits while middle 2 are for improvements.  Still take a long
  time to get to 2.00.01
  1.01.02 May 4 2005.  Will be putting in many changes - so saving stable version
  1.02.01 May 6 2005.  Lots of changes to try and make faster and more stable in
  branch and cut.
  1.02.02 May 19 2005.  Stuff for strong branching and some improvements to simplex
  1.03.01 May 24 2006.  Lots done but I can't remember what!
  1.03.03 June 13 2006.  For clean up after dual perturbation
  1.04.01 June 26 2007.  Lots of changes but I got lazy
  1.05.00 June 27 2007.  This is trunk so when gets to stable will be 1.5
  1.11.00 November 5 2009 (Guy Fawkes) - OSL factorization and better ordering
 */
#ifdef CILK_TEST
// -*- C++ -*-

/*
 * cilk-for.cilk
 *
 * Copyright (c) 2007-2008 Cilk Arts, Inc.  55 Cambridge Street,
 * Burlington, MA 01803.  Patents pending.  All rights reserved. You may
 * freely use the sample code to guide development of your own works,
 * provided that you reproduce this notice in any works you make that
 * use the sample code.  This sample code is provided "AS IS" without
 * warranty of any kind, either express or implied, including but not
 * limited to any implied warranty of non-infringement, merchantability
 * or fitness for a particular purpose.  In no event shall Cilk Arts,
 * Inc. be liable for any direct, indirect, special, or consequential
 * damages, or any other damages whatsoever, for any use of or reliance
 * on this sample code, including, without limitation, any lost
 * opportunity, lost profits, business interruption, loss of programs or
 * data, even if expressly advised of or otherwise aware of the
 * possibility of such damages, whether in an action of contract,
 * negligence, tort, or otherwise.
 *
 * This file demonstrates a Cilk++ for loop
 */

#include <cilk/cilk.h>
//#include <cilk/cilkview.h>
#include <cilk/reducer_max.h>
#include <cstdlib>
#include <iostream>

// cilk_for granularity.
#define CILK_FOR_GRAINSIZE 128

double dowork(double i)
{
  // Waste time:
  int j;
  double k = i;
  for (j = 0; j < 50000; ++j) {
    k += k / ((j + 1) * (k + 1));
  }

  return k;
}
static void doSomeWork(double *a, int low, int high)
{
  if (high - low > 300) {
    int mid = (high + low) >> 1;
    cilk_spawn doSomeWork(a, low, mid);
    doSomeWork(a, mid, high);
    cilk_sync;
  } else {
    for (int i = low; i < high; ++i) {
      a[i] = dowork(a[i]);
    }
  }
}

void cilkTest()
{
  unsigned int n = 10000;
  //cilk::cilkview cv;

  double *a = new double[n];

  for (unsigned int i = 0; i < n; i++) {
    // Populate A
    a[i] = (double)((i * i) % 1024 + 512) / 512;
  }

  std::cout << "Iterating over " << n << " integers" << std::endl;

  //cv.start();
#if 1
  //#pragma cilk grainsize = CILK_FOR_GRAINSIZE
  cilk_for(unsigned int i = 0; i < n; ++i)
  {
    a[i] = dowork(a[i]);
  }
#else
  doSomeWork(a, 0, n);
#endif
  int *which = new int[n];
  unsigned int n2 = n >> 1;
  for (int i = 0; i < n2; i++)
    which[i] = n - 2 * i;
  cilk::reducer_max_index< int, double > maximumIndex(-1, 0.0);
  cilk_for(unsigned int i = 0; i < n2; ++i)
  {
    int iWhich = which[i];
    maximumIndex.calc_max(iWhich, a[iWhich]);
  }
  int bestIndex = maximumIndex.get_index();
  int bestIndex2 = -1;
  double largest = 0.0;
  cilk_for(unsigned int i = 0; i < n2; ++i)
  {
    int iWhich = which[i];
    if (a[iWhich] > largest) {
      bestIndex2 = iWhich;
      largest = a[iWhich];
    }
  }
  assert(bestIndex == bestIndex2);
  //cv.stop();
  //cv.dump("cilk-for-results", false);

  //std::cout << cv.accumulated_milliseconds() / 1000.f << " seconds" << std::endl;

  exit(0);
}
#endif