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
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
|
#define FILENAME(line) FILENAME_FOR_EXCEPTIONS_CUDA("src/cuda-kernels/manual_awkward_Content_getitem_next_missing_jagged_getmaskstartstop.cu", line)
#include "awkward/kernels.h"
#include "standard_parallel_algorithms.h"
template <typename T>
__global__ void
awkward_MaskedArray_getitem_next_jagged_project_filter_mask(
T* index,
int64_t* filtered_index,
int64_t length) {
int64_t thread_id = blockIdx.x * blockDim.x + threadIdx.x;
if(thread_id < length) {
if (index[thread_id] >= 0) {
filtered_index[thread_id] = 1;
}
}
}
template <typename T>
__global__ void
awkward_MaskedArray_getitem_next_jagged_project_kernel(
T* index,
int64_t* prefixedsum_mask,
int64_t* starts_in,
int64_t* stops_in,
int64_t* starts_out,
int64_t* stops_out,
int64_t length) {
int64_t thread_id = blockIdx.x * blockDim.x + threadIdx.x;
if(thread_id < length) {
if (index[thread_id] >= 0) {
starts_out[prefixedsum_mask[thread_id] - 1] = starts_in[thread_id];
stops_out[prefixedsum_mask[thread_id] - 1] = stops_in[thread_id];
}
}
}
template <typename T>
ERROR awkward_MaskedArray_getitem_next_jagged_project(
T* index,
int64_t* starts_in,
int64_t* stops_in,
int64_t* starts_out,
int64_t* stops_out,
int64_t length) {
int64_t* res_temp;
int64_t* filtered_index;
dim3 blocks_per_grid = blocks(length);
dim3 threads_per_block = threads(length);
HANDLE_ERROR(cudaMalloc((void**)&res_temp, sizeof(int64_t) * length));
HANDLE_ERROR(cudaMalloc((void**)&filtered_index, sizeof(int64_t) * length));
HANDLE_ERROR(cudaMemset(filtered_index, 0, sizeof(int64_t) * length));
awkward_MaskedArray_getitem_next_jagged_project_filter_mask<<<
blocks_per_grid,
threads_per_block>>>(index, filtered_index, length);
exclusive_scan<int64_t, int64_t>(res_temp, filtered_index, length);
awkward_MaskedArray_getitem_next_jagged_project_kernel<<<blocks_per_grid,
threads_per_block>>>(
index,
res_temp,
starts_in,
stops_in,
starts_out,
stops_out,
length);
cudaDeviceSynchronize();
return success();
}
ERROR awkward_MaskedArray32_getitem_next_jagged_project(
int32_t* index,
int64_t* starts_in,
int64_t* stops_in,
int64_t* starts_out,
int64_t* stops_out,
int64_t length) {
return awkward_MaskedArray_getitem_next_jagged_project<int32_t>(
index,
starts_in,
stops_in,
starts_out,
stops_out,
length);
}
ERROR awkward_MaskedArrayU32_getitem_next_jagged_project(
uint32_t* index,
int64_t* starts_in,
int64_t* stops_in,
int64_t* starts_out,
int64_t* stops_out,
int64_t length) {
return awkward_MaskedArray_getitem_next_jagged_project<uint32_t>(
index,
starts_in,
stops_in,
starts_out,
stops_out,
length);
}
ERROR awkward_MaskedArray64_getitem_next_jagged_project(
int64_t* index,
int64_t* starts_in,
int64_t* stops_in,
int64_t* starts_out,
int64_t* stops_out,
int64_t length) {
return awkward_MaskedArray_getitem_next_jagged_project<int64_t>(
index,
starts_in,
stops_in,
starts_out,
stops_out,
length);
}
|
