Newer
Older
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
#include <cuda.h>
#include <cuda_runtime_api.h>
namespace memproc
{
template<>
bool dealloc<MemType::GPU>(void *&array, size_t &allocated_size)
{
if(allocated_size > 0)
{
cudaFree(array);
allocated_size = 0;
}
return true;
}
template<>
bool dealloc<MemType::GPU>(void *&array)
{
cudaFree(array);
return true;
}
template <>
bool alloc<MemType::GPU>(void *&array, const size_t new_size)
{
cudaMalloc ( (void **)&array, new_size);
cudaMemset(array, 0, new_size);
return true;
}
template <>
bool realloc<MemType::GPU>(void *&array, size_t &allocated_size, const size_t new_size)
{
if(new_size > allocated_size)
{
if(allocated_size > 0) dealloc<MemType::GPU>(array, allocated_size);
allocated_size = new_size;
cudaMalloc ( (void **)&array, new_size);
cudaMemset(array, 0, new_size);
}
return true;
}
template <>
bool memcopy<MemType::GPU, MemType::CPU>(void *dst, const void* src, const size_t copy_elem_size)
{
cudaMemcpy ( dst, src, copy_elem_size, cudaMemcpyHostToDevice);
return true;
}
template <>
bool memcopy<MemType::CPU, MemType::GPU>(void *dst, const void* src, const size_t copy_elem_size)
{
cudaMemcpy ( dst, src, copy_elem_size, cudaMemcpyDeviceToHost);
return true;
}
template <>
bool memcopy<MemType::GPU, MemType::GPU>(void *dst, const void* src, const size_t copy_elem_size)
{
cudaMemcpy ( dst, src, copy_elem_size, cudaMemcpyDeviceToDevice);
return true;
}
}