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
|
// BSD 3-Clause License; see https://github.com/scikit-hep/awkward-1.0/blob/main/LICENSE
#define FILENAME(line) \
FILENAME_FOR_EXCEPTIONS_CUDA( \
"src/cpu-kernels/" \
"awkward_ByteMaskedArray_reduce_next_nonlocal_nextshifts_64.cpp", \
line)
#include "standard_parallel_algorithms.h"
#include "awkward/kernels.h"
__global__ void
awkward_ByteMaskedArray_reduce_next_nonlocal_nextshifts_64_filter_mask(
int8_t* filtered_mask_k,
int8_t* filtered_mask_nullsum,
const int8_t* mask,
bool valid_when,
int64_t length) {
int64_t thread_id = blockIdx.x * blockDim.x + threadIdx.x;
if (thread_id < length) {
if ((mask[thread_id] != 0) == (valid_when != 0)) {
filtered_mask_k[thread_id] = 1;
} else {
filtered_mask_nullsum[thread_id] = 1;
}
}
}
__global__ void
awkward_ByteMaskedArray_reduce_next_nonlocal_nextshifts_64_kernel(
int64_t* nextshifts,
const int8_t* mask,
bool valid_when,
int64_t length,
int64_t* prefixed_mask_k,
int64_t* prefixed_mask_nullsum) {
int64_t thread_id = blockIdx.x * blockDim.x + threadIdx.x;
if (thread_id < length) {
if ((mask[thread_id] != 0) == (valid_when != 0)) {
nextshifts[prefixed_mask_k[thread_id] - 1] =
prefixed_mask_nullsum[thread_id] - 1;
}
}
}
ERROR
awkward_ByteMaskedArray_reduce_next_nonlocal_nextshifts_64(int64_t* nextshifts,
const int8_t* mask,
int64_t length,
bool valid_when) {
dim3 blocks_per_grid = blocks(length);
dim3 threads_per_block = threads(length);
int8_t* filtered_mask_k;
int8_t* filtered_mask_nullsum;
HANDLE_ERROR(cudaMalloc((void**)&filtered_mask_k, sizeof(int8_t) * length));
HANDLE_ERROR(
cudaMalloc((void**)&filtered_mask_nullsum, sizeof(int8_t) * length));
HANDLE_ERROR(cudaMemset(filtered_mask_k, 0, sizeof(int8_t) * length));
HANDLE_ERROR(cudaMemset(filtered_mask_nullsum, 0, sizeof(int8_t) * length));
awkward_ByteMaskedArray_reduce_next_nonlocal_nextshifts_64_filter_mask<<<
blocks_per_grid,
threads_per_block>>>(
filtered_mask_k, filtered_mask_nullsum, mask, valid_when, length);
int64_t* prefixed_mask_k;
HANDLE_ERROR(cudaMalloc((void**)&prefixed_mask_k, sizeof(int64_t) * length));
int64_t* prefixed_mask_nullsum;
HANDLE_ERROR(
cudaMalloc((void**)&prefixed_mask_nullsum, sizeof(int64_t) * length));
exclusive_scan(prefixed_mask_k, filtered_mask_k, length);
exclusive_scan(prefixed_mask_nullsum, filtered_mask_nullsum, length);
awkward_ByteMaskedArray_reduce_next_nonlocal_nextshifts_64_kernel<<<
blocks_per_grid,
threads_per_block>>>(
nextshifts, mask, valid_when, length, prefixed_mask_k, prefixed_mask_nullsum);
cudaDeviceSynchronize();
return success();
}
|
