/*
Written by Holguer A Becerra
UBC Student
Master of Biomedical Engineering 
Course: ECE 328 
Prof. Guy Lemiuex.
Year 2015-Winter Term 1
*/


#include <stdio.h>
#include <time.h>
#include <stdlib.h>
#include <windows.h>
 

#define INT 0 
#define FLOAT 1 
#define DOUBLE 1 

#define TYPE_VAR INT

#if TYPE_VAR == INT
	int * matrix1;
	int * matrix2;
	int * matrix3;
	int sum;
#elif TYPE_VAR == FLOAT
	float * matrix1;
	float * matrix2;
	float * matrix3;
	float sum;
#elif TYPE_VAR == DOUBLE
	double * matrix1;
	double * matrix2;
	double * matrix3;
	double sum;
#endif 
 
 
double PCFreq = 0.0;
__int64 CounterStart = 0;
double performance_counter=0.0;

void StartCounter();
double GetCounter();

int main(int argc, char *argv[])
{
	int N,rows,columns;
	int c,d,k;
	if(argc<=1){
		printf("too few arguments");
		return 0;
	}
	
	// the first argument has the number of rows, in this case the number of columns are the same as the number of rows.
	sscanf(argv[1],"%d",&N);
	

	/* initialize random seed: */
	srand (time(NULL));
	
	StartCounter();
	
  // Initialize Matrices	
	#if TYPE_VAR == INT
		matrix1=(int *)malloc(N*N*sizeof(int));
		matrix2=(int *)malloc(N*N*sizeof(int));
		matrix3=(int *)malloc(N*N*sizeof(int));
	#elif TYPE_VAR == FLOAT
		matrix1=(float *)malloc(N*N*sizeof(float));
		matrix2=(float *)malloc(N*N*sizeof(float));
		matrix3=(float *)malloc(N*N*sizeof(float));
	#elif TYPE_VAR == DOUBLE
		matrix1=(double *)malloc(N*N*sizeof(double));
		matrix2=(double *)malloc(N*N*sizeof(double));
		matrix3=(double *)malloc(N*N*sizeof(double));
	#endif 

	printf("Initializing matrices\n");
	for (rows= 0; rows < N*N; rows++){
		#if TYPE_VAR == INT
			*(matrix1+rows)=rand() % 1000 + 1;
			*(matrix2+rows)=rand() % 1000 + 7;
			//printf("1[%d,%d]=%d\n",rows/N,rows,*(matrix1+rows));
			//printf("2[%d,%d]=%d\n",rows/N,rows,*(matrix2+rows));
		#elif TYPE_VAR == FLOAT
			*(matrix1+rows)=rand() % 1000 + 1.5;
			*(matrix2+rows)=rand() % 1000 + 7.07;
			//printf("1[%d,%d]=%f\n",rows/N,rows,*(matrix1+rows));
			//printf("2[%d,%d]=%f\n",rows/N,rows,*(matrix2+rows));
		#elif TYPE_VAR == DOUBLE
			*(matrix1+rows)=rand() % 1000 + 1.5;
			*(matrix2+rows)=rand() % 1000 + 7.07;			
			//printf("1[%d,%d]=%f\n",rows/N,rows,*(matrix1+rows));
			//printf("2[%d,%d]=%f\n",rows/N,rows,*(matrix2+rows));
		#endif 
		
	}
  
 
  
  // Performance  Start
  StartCounter();
  
  
  
  
  
  
  
  
  
  
	// Improved 
	int ho=0;
    for (c = 0; c < N; c++) {
	  ho=c*N;
      for (k = 0; k < N; k++) {
        for (d = 0; d < N; d++) {
			*(matrix3+ho+d) += (*(matrix1+ho+k))*(*(matrix2+k*N+d));
            //printf("C(%d,%d)=A(%d,%d)*B(%d,%d);\n",ho,d,ho,k,k*N,d);
        }
 
      }
    }
	
	
	
	
	
	
	
	
	
	
	
	

  // Performance  Finish and Measure
  performance_counter=GetCounter();
  
  

 printf("\n Improved Mutiplication time: %f ms\n",performance_counter);	

  // free memory
  
  free(matrix1);
  free(matrix2);
  
  

 
  return 0;
}




void StartCounter()
{
    LARGE_INTEGER li;
    if(!QueryPerformanceFrequency(&li))
	{
	printf("QueryPerformanceFrequency failed!\n");
	}

	PCFreq = (double)li.QuadPart/1000.0;

    QueryPerformanceCounter(&li);
    CounterStart = li.QuadPart;
}

double GetCounter()
{
    LARGE_INTEGER li;
    QueryPerformanceCounter(&li);
    return (double)(li.QuadPart-CounterStart)/PCFreq;
}