#include <iostream>
#include <memory>
#include <omp.h>
#include <random>
#include "brick.h"
#include "bricksetup.h"
#include "multiarray.h"
#include "brickcompare.h"
#include <immintrin.h>

Include dependency graph for main.cpp:

Macros
#define	VSVEC "AVX2"

#define	VFOLD 2,2

#define	N 256

#define	TILE 8

#define	GZ TILE

#define	PADDING 8

#define	STRIDE (N + 2 * (GZ + PADDING))

#define	STRIDEG (N + 2 * GZ)

#define	NB (N / TILE)

#define	GB (GZ / TILE)

#define	STRIDEB ((N + 2 * GZ) / TILE)

#define	BDIM TILE,TILE,TILE

#define	TOT_TIME 5

#define	_PARFOR _Pragma("omp parallel for collapse(2)")

#define	_TILEFOR

#define	bIn(i, j, k) arr_in[k][j][i]

#define	bOut(i, j, k) arr_out[k][j][i]

Functions
template<typename T >
double	time_func (T func)

void	d3pt7 ()

int	main ()

Variables
bElem *	coeff

Macro Definition Documentation

◆ _PARFOR

#define _PARFOR _Pragma("omp parallel for collapse(2)")

◆ _TILEFOR

#define _TILEFOR

Value:

_Pragma("omp parallel for collapse(2)") \
for (long tk = PADDING; tk < PADDING + STRIDEG; tk += TILE) \
for (long tj = PADDING; tj < PADDING + STRIDEG; tj += TILE) \
for (long ti = PADDING; ti < PADDING + STRIDEG; ti += TILE) \
for (long k = tk; k < tk + TILE; ++k) \
for (long j = tj; j < tj + TILE; ++j) \
_Pragma("omp simd") \
for (long i = ti; i < ti + TILE; ++i)

◆ BDIM

#define BDIM TILE,TILE,TILE

◆ bIn

#define bIn	(	i,
		j,
		k
	)	arr_in[k][j][i]

◆ bOut

#define bOut	(	i,
		j,
		k
	)	arr_out[k][j][i]

◆ GB

#define GB (GZ / TILE)

◆ GZ

#define GZ TILE

◆ N

#define N 256

◆ NB

#define NB (N / TILE)

◆ PADDING

#define PADDING 8

◆ STRIDE

#define STRIDE (N + 2 * (GZ + PADDING))

◆ STRIDEB

#define STRIDEB ((N + 2 * GZ) / TILE)

◆ STRIDEG

#define STRIDEG (N + 2 * GZ)

◆ TILE

#define TILE 8

◆ TOT_TIME

#define TOT_TIME 5

◆ VFOLD

#define VFOLD 2,2

◆ VSVEC

#define VSVEC "AVX2"

Function Documentation

◆ d3pt7()

void d3pt7 ( )

Here is the call graph for this function:

◆ main()

int main ( )