-
-
Notifications
You must be signed in to change notification settings - Fork 202
/
xaxpy.opencl
113 lines (98 loc) · 4.4 KB
/
xaxpy.opencl
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
// =================================================================================================
// This file is part of the CLBlast project. The project is licensed under Apache Version 2.0. This
// project loosely follows the Google C++ styleguide and uses a tab-size of two spaces and a max-
// width of 100 characters per line.
//
// Author(s):
// Cedric Nugteren <www.cedricnugteren.nl>
//
// This file contains the Xaxpy kernel. It contains one fast vectorized version in case of unit
// strides (incx=incy=1) and no offsets (offx=offy=0). Another version is more general, but doesn't
// support vector data-types. The general version has a batched implementation as well.
//
// This kernel uses the level-1 BLAS common tuning parameters.
//
// =================================================================================================
// Enables loading of this file using the C++ pre-processor's #include (C++11 standard raw string
// literal). Comment-out this line for syntax-highlighting when developing.
R"(
// =================================================================================================
// Full version of the kernel with offsets and strided accesses
#if RELAX_WORKGROUP_SIZE == 1
__kernel
#else
__kernel __attribute__((reqd_work_group_size(WGS, 1, 1)))
#endif
void Xaxpy(const int n, const real_arg arg_alpha,
const __global real* restrict xgm, const int x_offset, const int x_inc,
__global real* ygm, const int y_offset, const int y_inc) {
const real alpha = GetRealArg(arg_alpha);
// Loops over the work that needs to be done (allows for an arbitrary number of threads)
for (int id = get_global_id(0); id < n; id += get_global_size(0)) {
real xvalue = xgm[id*x_inc + x_offset];
MultiplyAdd(ygm[id*y_inc + y_offset], alpha, xvalue);
}
}
// Faster version of the kernel without offsets and strided accesses but with if-statement. Also
// assumes that 'n' is dividable by 'VW' and 'WPT'.
#if RELAX_WORKGROUP_SIZE == 1
__kernel
#else
__kernel __attribute__((reqd_work_group_size(WGS, 1, 1)))
#endif
void XaxpyFaster(const int n, const real_arg arg_alpha,
const __global realV* restrict xgm,
__global realV* ygm) {
const real alpha = GetRealArg(arg_alpha);
const int num_usefull_threads = n / (VW * WPT);
if (get_global_id(0) < num_usefull_threads) {
#pragma unroll
for (int _w = 0; _w < WPT; _w += 1) {
const int id = _w*num_usefull_threads + get_global_id(0);
realV xvalue = xgm[id];
realV yvalue = ygm[id];
ygm[id] = MultiplyAddVector(yvalue, alpha, xvalue);
}
}
}
// Faster version of the kernel without offsets and strided accesses. Also assumes that 'n' is
// dividable by 'VW', 'WGS' and 'WPT'.
#if RELAX_WORKGROUP_SIZE == 1
__kernel
#else
__kernel __attribute__((reqd_work_group_size(WGS, 1, 1)))
#endif
void XaxpyFastest(const int n, const real_arg arg_alpha,
const __global realV* restrict xgm,
__global realV* ygm) {
const real alpha = GetRealArg(arg_alpha);
#pragma unroll
for (int _w = 0; _w < WPT; _w += 1) {
const int id = _w*get_global_size(0) + get_global_id(0);
realV xvalue = xgm[id];
realV yvalue = ygm[id];
ygm[id] = MultiplyAddVector(yvalue, alpha, xvalue);
}
}
// =================================================================================================
// Full version of the kernel with offsets and strided accesses: batched version
#if RELAX_WORKGROUP_SIZE == 1
__kernel
#else
__kernel __attribute__((reqd_work_group_size(WGS, 1, 1)))
#endif
void XaxpyBatched(const int n, const __constant real_arg* arg_alphas,
const __global real* restrict xgm, const __constant int* x_offsets, const int x_inc,
__global real* ygm, const __constant int* y_offsets, const int y_inc) {
const int batch = get_group_id(1);
const real alpha = GetRealArg(arg_alphas[batch]);
// Loops over the work that needs to be done (allows for an arbitrary number of threads)
for (int id = get_global_id(0); id < n; id += get_global_size(0)) {
real xvalue = xgm[id*x_inc + x_offsets[batch]];
MultiplyAdd(ygm[id*y_inc + y_offsets[batch]], alpha, xvalue);
}
}
// =================================================================================================
// End of the C++11 raw string literal
)"
// =================================================================================================