defs.h 2.64 KB
Newer Older
1 2 3
#ifndef BIOEM_DEFS_H
#define BIOEM_DEFS_H

David Rohr's avatar
David Rohr committed
4
//#define BIOEM_USE_DOUBLE
5

6
#ifndef BIOEM_USE_DOUBLE
7
typedef float myfloat_t;
8 9 10 11
#define myfftw_malloc fftwf_malloc
#define myfftw_free fftwf_free
#define myfftw_destroy_plan fftwf_destroy_plan
#define myfftw_execute fftwf_execute
12
#define myfftw_execute_dft fftwf_execute_dft
13 14
#define myfftw_execute_dft_r2c fftwf_execute_dft_r2c
#define myfftw_execute_dft_c2r fftwf_execute_dft_c2r
15
#define myfftw_plan_dft_2d fftwf_plan_dft_2d
16 17
#define myfftw_plan_dft_r2c_2d fftwf_plan_dft_r2c_2d
#define myfftw_plan_dft_c2r_2d fftwf_plan_dft_c2r_2d
18
#define myfftw_plan fftwf_plan
19
#define myfftw_cleanup fftwf_cleanup
David Rohr's avatar
David Rohr committed
20
#define MY_CUFFT_C2R CUFFT_C2R
21 22
#define mycufftExecC2R cufftExecC2R
#define mycuComplex_t cuComplex
David Rohr's avatar
David Rohr committed
23
#define MY_MPI_FLOAT MPI_FLOAT
24 25 26 27 28 29
#else
typedef double myfloat_t;
#define myfftw_malloc fftw_malloc
#define myfftw_free fftw_free
#define myfftw_destroy_plan fftw_destroy_plan
#define myfftw_execute fftw_execute
30
#define myfftw_execute_dft fftw_execute_dft
31 32
#define myfftw_execute_dft_r2c fftw_execute_dft_r2c
#define myfftw_execute_dft_c2r fftw_execute_dft_c2r
33
#define myfftw_plan_dft_2d fftw_plan_dft_2d
34 35
#define myfftw_plan_dft_r2c_2d fftw_plan_dft_r2c_2d
#define myfftw_plan_dft_c2r_2d fftw_plan_dft_c2r_2d
36
#define myfftw_plan fftw_plan
37
#define myfftw_cleanup fftw_cleanup
38 39
#define mycufftExecC2R cufftExecZ2D
#define mycuComplex_t cuDoubleComplex
David Rohr's avatar
David Rohr committed
40
#define MY_CUFFT_C2R CUFFT_Z2D
David Rohr's avatar
David Rohr committed
41
#define MY_MPI_FLOAT MPI_DOUBLE
42 43
#endif
typedef myfloat_t mycomplex_t[2];
44 45 46 47 48

#define BIOEM_FLOAT_3_PHYSICAL_SIZE 3	//Possible set to 4 for GPU

struct myfloat3_t
{
49
	myfloat_t pos[BIOEM_FLOAT_3_PHYSICAL_SIZE];
50 51 52 53 54 55 56 57 58
};

#ifdef BIOEM_GPUCODE
#define myThreadIdxX threadIdx.x
#define myThreadIdxY threadIdx.y
#define myBlockDimX blockDim.x
#define myBlockDimY blockDim.y
#define myBlockIdxX blockIdx.x
#define myBlockIdxY blockIdx.y
59
#define myGridDimX gridDim.x
60 61 62 63 64 65 66 67 68 69 70 71 72 73
#else
#define __device__
#define __host__
#define myThreadIdxX 0
#define myThreadIdxY 0
#define myBlockDimX 1
#define myBlockDimY 1
#define myBlockIdxX 0
#define myBlockIdxY 0
#endif

#define CUDA_THREAD_COUNT 256
#define CUDA_BLOCK_COUNT 1024 * 16
#define CUDA_MAX_SHIFT_REDUCE 1024
74
#define CUDA_FFTS_AT_ONCE 1024
75
//#define BIOEM_USE_NVTX
76

77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98
static inline void* mallocchk(size_t size)
{
	void* ptr = malloc(size);
	if (ptr == 0)
	{
		std::cout << "Memory allocation error\n";
		exit(1);
	}
	return(ptr);
}

static inline void* reallocchk(void* oldptr, size_t size)
{
	void* ptr = realloc(oldptr, size);
	if (ptr == 0)
	{
		std::cout << "Memory allocation error\n";
		exit(1);
	}
	return(ptr);
}

99 100 101 102 103
#ifndef WITH_OPENMP
#define omp_get_max_threads() 1
#define omp_get_thread_num() 0
#endif

104 105 106
extern int mpi_rank;
extern int mpi_size;

107
#endif