forked from ROCm/pytorch
-
Notifications
You must be signed in to change notification settings - Fork 0
/
UnaryOpsKernel.cpp
245 lines (222 loc) · 8.74 KB
/
UnaryOpsKernel.cpp
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
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
#include "ATen/native/cpu/UnaryOpsKernel.h"
#include <cmath>
#include <type_traits>
#include "ATen/Config.h"
#include "ATen/Dispatch.h"
#include "ATen/CPUGenerator.h"
#include "ATen/CheckGenerator.h"
#include "ATen/Generator.h"
#include "ATen/cpu/vml.h"
#include "ATen/CPUApplyUtils.h"
#include "ATen/native/DispatchStub.h"
#include "ATen/native/Distributions.h"
#ifdef __AVX2__
#include "ATen/native/cpu/avx_mathfun.h"
#endif
#if AT_MKL_ENABLED()
#include <mkl.h>
#endif
#include "TH/THGenerator.hpp"
#include "TH/THRandom.h"
namespace at { namespace native {
namespace {
using namespace vec256;
template <typename scalar_t>
static int64_t _sigmoid(scalar_t* x, scalar_t* y, int64_t size);
// This should be a temporary solution until we understand why SLEEF is slower
// for sigmoid
template <>
int64_t _sigmoid(float* x, float* y, int64_t size) {
using Vec = Vec256<float>;
int64_t i = 0;
for (; i < size - (size % (2 * Vec::size)); i += 2 * Vec::size) {
Vec ret = Vec::loadu(y + i);
Vec ret2 = Vec::loadu(y + i + Vec::size);
ret = ret.neg();
ret2 = ret2.neg();
#if defined(__AVX2__) && !defined(_MSC_VER)
ret = exp256_ps(ret);
ret2 = exp256_ps(ret2);
#else
ret = ret.exp();
ret2 = ret2.exp();
#endif
ret = Vec((float)(1)) + ret;
ret2 = Vec((float)(1)) + ret2;
ret = ret.reciprocal();
ret2 = ret2.reciprocal();
ret.store(x + i);
ret2.store(x + i + Vec::size);
}
return i;
}
template <>
int64_t _sigmoid(double* x, double* y, int64_t size) {
using Vec = Vec256<double>;
int64_t i = 0;
for (; i < size - (size % (2 * Vec::size)); i += 2 * Vec::size) {
Vec ret = Vec::loadu(y + i);
Vec ret2 = Vec::loadu(y + i + Vec::size);
ret = ret.neg();
ret2 = ret2.neg();
ret = ret.exp();
ret2 = ret2.exp();
ret = Vec((double)(1)) + ret;
ret2 = Vec((double)(1)) + ret2;
ret = ret.reciprocal();
ret2 = ret2.reciprocal();
ret.store(x + i);
ret2.store(x + i + Vec::size);
}
return i;
}
static void sigmoid_kernel(Tensor& result, const Tensor& self) {
AT_DISPATCH_FLOATING_TYPES(self.type(), "sigmoid", [&] {
using Vec = Vec256<scalar_t>;
CPU_tensor_parallel_kernel_apply2<scalar_t, scalar_t>(
result,
self,
[](int64_t size,
scalar_t* x,
scalar_t* y,
int64_t stridex,
int64_t stridey) {
int64_t i = 0;
if (stridex == 1 && stridey == 1) {
i = _sigmoid(x, y, size);
}
for (; i < size; i += Vec::size) {
scalar_t buffer[Vec::size];
int64_t width = Vec::size;
width = std::min(width, size - i);
for (int64_t j = 0; j < width; j++) {
buffer[j] = y[stridey * (i + j)];
}
Vec ret = Vec::loadu(buffer);
ret = Vec((scalar_t)(0)) - ret;
ret = ret.exp();
ret = Vec((scalar_t)(1)) + ret;
ret = ret.reciprocal();
ret.store(buffer);
for (int64_t j = 0; j < width; j++)
x[stridex * (i + j)] = buffer[j];
}
});
});
}
#if !AT_MKL_ENABLED()
void bernoulli_mkl_kernel(Tensor &output, const double p, Generator* gen) {
// Use AT_ASSERTM because this should never be reached, and AT_ASSERTM tells
// users to report this as a bug.
AT_ASSERTM(false, "ATen not compiled with MKL");
}
#else
void bernoulli_mkl_kernel(Tensor &self, const double p, Generator* gen) {
THGenerator* generator = get_generator(gen);
int64_t seed;
{
std::lock_guard<std::mutex> lock(generator->mutex);
seed = THRandom_random(generator);
}
int64_t n = self.numel();
bool contig = self.is_contiguous();
AT_DISPATCH_ALL_TYPES(self.type(), "bernoulli_scalar_cpu_", [&] {
at::Tensor tmp_int_tensor;
if (std::is_same<scalar_t, int>::value && contig) {
tmp_int_tensor = self;
} else {
tmp_int_tensor = at::empty(self.sizes(), self.options().dtype(at::kInt));
}
scalar_t *self_ptr = self.data<scalar_t>();
int *sample_int_ptr = tmp_int_tensor.data<int>();
auto sample = [&](int64_t begin, int64_t end) {
int64_t len = end - begin;
if (len > 0) {
VSLStreamStatePtr stream;
vslNewStream(&stream, VSL_BRNG_MCG31, seed);
vslSkipAheadStream(stream, begin);
viRngBernoulli(VSL_RNG_METHOD_BERNOULLI_ICDF, stream, len,
sample_int_ptr + begin, p);
vslDeleteStream(&stream);
// vectorized copy if using buffer and contiguous, i.e., being non-int
// type and contiguous
if (!std::is_same<scalar_t, int>::value && contig) {
scalar_t *self_seg = self_ptr + begin;
int* tmp_seg = sample_int_ptr + begin;
at::vec256::convert<int, scalar_t>(tmp_seg, self_seg, len);
}
}
};
parallel_for(0, n, /* grain_size= */ 800, sample);
// copy_ if using buffer and non contiguous
if (!contig) {
self.copy_(tmp_int_tensor);
}
});
}
#endif
#define IMPLEMENT_FLOAT_KERNEL(dispatchtypes, op) \
static void op##_kernel(Tensor& result, const Tensor& self) { \
checkBackend(#op, {result}, Backend::CPU); \
AT_DISPATCH_##dispatchtypes##_TYPES(self.type(), #op, [&] { \
if (self.is_contiguous() && result.is_contiguous()) { \
vml::v##op( \
result.data<scalar_t>(), self.data<scalar_t>(), self.numel()); \
\
} else { \
static constexpr int64_t WIDTH = 131072 / sizeof(scalar_t); \
CPU_tensor_parallel_kernel_apply2<scalar_t, scalar_t>( \
result, \
self, \
[](int64_t size, \
scalar_t* x, \
scalar_t* y, \
int64_t stridex, \
int64_t stridey) { \
if (stridex == 1 && stridey == 1) { \
vml::v##op(x, y, size); \
} else { \
for (int64_t i = 0; i < size; i += WIDTH) { \
scalar_t buffer[WIDTH]; \
int64_t width = WIDTH; \
width = std::min(width, size - i); \
for (int64_t j = 0; j < width; j++) \
buffer[j] = y[stridey * (i + j)]; \
vml::v##op(buffer, buffer, width); \
for (int64_t j = 0; j < width; j++) \
x[stridex * (i + j)] = buffer[j]; \
} \
} \
}); \
} \
}); \
} \
REGISTER_DISPATCH(op##Impl, &op##_kernel)
} // anonymous namespace
REGISTER_DISPATCH(sigmoidImpl, &sigmoid_kernel)
REGISTER_DISPATCH(bernoulli_mkl_stub, &bernoulli_mkl_kernel);
// IMPLEMENT_FLOAT_KERNEL(ALL, abs)
IMPLEMENT_FLOAT_KERNEL(FLOATING, acos)
IMPLEMENT_FLOAT_KERNEL(FLOATING, asin)
IMPLEMENT_FLOAT_KERNEL(FLOATING, atan)
IMPLEMENT_FLOAT_KERNEL(FLOATING, ceil)
IMPLEMENT_FLOAT_KERNEL(FLOATING, cos)
// IMPLEMENT_FLOAT_KERNEL(FLOATING, cosh)
IMPLEMENT_FLOAT_KERNEL(FLOATING, erf)
IMPLEMENT_FLOAT_KERNEL(FLOATING, erfc)
IMPLEMENT_FLOAT_KERNEL(FLOATING, exp)
IMPLEMENT_FLOAT_KERNEL(FLOATING, expm1)
IMPLEMENT_FLOAT_KERNEL(FLOATING, floor)
IMPLEMENT_FLOAT_KERNEL(FLOATING, log)
IMPLEMENT_FLOAT_KERNEL(FLOATING, log10)
IMPLEMENT_FLOAT_KERNEL(FLOATING, log1p)
IMPLEMENT_FLOAT_KERNEL(FLOATING, log2)
IMPLEMENT_FLOAT_KERNEL(FLOATING, round)
IMPLEMENT_FLOAT_KERNEL(FLOATING, rsqrt)
IMPLEMENT_FLOAT_KERNEL(FLOATING, sin)
// IMPLEMENT_FLOAT_KERNEL(FLOATING, sinh)
IMPLEMENT_FLOAT_KERNEL(FLOATING, sqrt)
IMPLEMENT_FLOAT_KERNEL(FLOATING, tan)
IMPLEMENT_FLOAT_KERNEL(FLOATING, tanh)
IMPLEMENT_FLOAT_KERNEL(FLOATING, trunc)
}} // namespace at::native