darknet-sys 0.3.1

-sys crate for Rust darknet wrapper
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106

#ifndef DARKCUDA_H
#define DARKCUDA_H
#include "darknet.h"

#ifdef __cplusplus
extern "C" {
#endif


extern int cuda_debug_sync;
extern int gpu_index;
#ifdef __cplusplus
}
#endif // __cplusplus

#ifdef GPU

#define BLOCK 512
#define FULL_MASK 0xffffffff
#define WARP_SIZE 32
#define BLOCK_TRANSPOSE32 256

#include <cuda.h>
#include <cuda_runtime.h>
#include <curand.h>
#include <cublas_v2.h>
#include <cuda_runtime_api.h>
//#include <driver_types.h>

#ifdef CUDNN
#include <cudnn.h>
#endif // CUDNN

#ifndef __DATE__
#define __DATE__
#endif

#ifndef __TIME__
#define __TIME__
#endif

#ifndef __FUNCTION__
#define __FUNCTION__
#endif

#ifndef __LINE__
#define __LINE__ 0
#endif

#ifndef __FILE__
#define __FILE__
#endif

#ifdef __cplusplus
extern "C" {
#endif // __cplusplus
    void check_error(cudaError_t status);
    void check_error_extended(cudaError_t status, const char *file, int line, const char *date_time);
    void cublas_check_error_extended(cublasStatus_t status, const char *file, int line, const char *date_time);
#define CHECK_CUDA(X) check_error_extended(X, __FILE__ " : " __FUNCTION__, __LINE__,  __DATE__ " - " __TIME__ );
#define CHECK_CUBLAS(X) cublas_check_error_extended(X, __FILE__ " : " __FUNCTION__, __LINE__,  __DATE__ " - " __TIME__ );

    cublasHandle_t blas_handle();
    void free_pinned_memory();
    void pre_allocate_pinned_memory(size_t size);
    float *cuda_make_array_pinned_preallocated(float *x, size_t n);
    float *cuda_make_array_pinned(float *x, size_t n);
    float *cuda_make_array(float *x, size_t n);
    void **cuda_make_array_pointers(void **x, size_t n);
    int *cuda_make_int_array(size_t n);
	int *cuda_make_int_array_new_api(int *x, size_t n);
    void cuda_push_array(float *x_gpu, float *x, size_t n);
    //LIB_API void cuda_pull_array(float *x_gpu, float *x, size_t n);
    //LIB_API void cuda_set_device(int n);
    int cuda_get_device();
    void cuda_free_host(float *x_cpu);
    void cuda_free(float *x_gpu);
    void cuda_random(float *x_gpu, size_t n);
    float cuda_compare(float *x_gpu, float *x, size_t n, char *s);
    dim3 cuda_gridsize(size_t n);
    cudaStream_t get_cuda_stream();
    //cudaStream_t get_cuda_memcpy_stream();
    int get_number_of_blocks(int array_size, int block_size);
    int get_gpu_compute_capability(int i, char *device_name);
    void show_cuda_cudnn_info();

    cudaStream_t switch_stream(int i);
    void wait_stream(int i);
    void reset_wait_stream_events();

#ifdef CUDNN
cudnnHandle_t cudnn_handle();
enum {cudnn_fastest, cudnn_smallest, cudnn_specify};

void cudnn_check_error_extended(cudnnStatus_t status, const char *file, int line, const char *date_time);
#define CHECK_CUDNN(X) cudnn_check_error_extended(X, __FILE__ " : " __FUNCTION__, __LINE__,  __DATE__ " - " __TIME__ );
#endif

#ifdef __cplusplus
}
#endif // __cplusplus

#else // GPU
//LIB_API void cuda_set_device(int n);
#endif // GPU
#endif // DARKCUDA_H