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
|
/*
* Copyright (c) 2017-2022 Arm Limited.
*
* SPDX-License-Identifier: MIT
*
* Permission is hereby granted, free of charge, to any person obtaining a copy
* of this software and associated documentation files (the "Software"), to
* deal in the Software without restriction, including without limitation the
* rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
* sell copies of the Software, and to permit persons to whom the Software is
* furnished to do so, subject to the following conditions:
*
* The above copyright notice and this permission notice shall be included in all
* copies or substantial portions of the Software.
*
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
* AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
* LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
* OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
* SOFTWARE.
*/
#ifndef ARM_COMPUTE_CPU_GEMMLOWP_REDUCTION_KERNEL_H
#define ARM_COMPUTE_CPU_GEMMLOWP_REDUCTION_KERNEL_H
#include "src/core/common/Macros.h"
#include "src/cpu/ICpuKernel.h"
namespace arm_compute
{
// Forward declarations
struct GEMMLowpReductionKernelInfo;
namespace cpu
{
namespace kernels
{
/** Kernel used to compute the row-vectors of sums of all the entries in each row of Matrix A.
*
* @note This stage is needed to handle the offset of matrix product
* https://github.com/google/gemmlowp/blob/master/doc/low-precision.md
*/
class CpuGemmLowpMatrixAReductionKernel : public ICpuKernel<CpuGemmLowpMatrixAReductionKernel>
{
public:
/** Default constructor */
CpuGemmLowpMatrixAReductionKernel() = default;
ARM_COMPUTE_DISALLOW_COPY_ALLOW_MOVE(CpuGemmLowpMatrixAReductionKernel);
/** Initialise the kernel's input and output.
*
* @param[in] src Input tensor. Data type supported: QASYMM8/QASYMM8_SIGNED/QSYMM8/QSYMM8_PER_CHANNEL
* @param[out] dst Output row-vector of sums of all the entries in each row of mtx_a. Data type supported: S32
* @param[in] info Kernel metadata:
* - k (num_mtx_a_cols) Number of matrix A columns
* - is_reshaped (is_interleaved4x4) True if the matrix A has been interleaved4x4
* - scalar Scalar value to multiply each reduced row by.
* - mul_byscalar True if each reduced column must be multiplied by a scalar value.
*/
void configure(const ITensorInfo *src, ITensorInfo *dst, const GEMMLowpReductionKernelInfo &info);
/** Static function to check if given info will lead to a valid configuration
*
* Similar to CpuGemmLowpMatrixAReductionKernel::configure()
*
* @return a status
*/
static Status validate(const ITensorInfo *src, const ITensorInfo *dst, const GEMMLowpReductionKernelInfo &info);
// Inherited methods overridden:
void run_op(ITensorPack &tensors, const Window &window, const ThreadInfo &info) override;
const char *name() const override;
private:
/** Execution of the reduction kernel specialized on the input type
*
* @param[in] src Input tensor
* @param[in] dst Output tensor
* @param[in] window Execution window
*/
template <typename T>
void run_internal(const ITensor *src, ITensor *dst, const Window &window);
/** Common signature for all reduction functions
*
* @param[in] src Input tensor
* @param[out] dst Output tensor
* @param[in] window Region on which to execute the kernel. (Must be a valid region of the window returned by window()).
*/
using CpuGemmLowpMatrixAReductionKernelPtr = void (CpuGemmLowpMatrixAReductionKernel::*)(const ITensor *src,
ITensor *dst,
const Window &window);
CpuGemmLowpMatrixAReductionKernelPtr _func{nullptr};
int32_t _k{0};
int32_t _scalar{0};
bool _mul_by_scalar{false};
};
/** Kernel used to compute the row-vectors of sums of all the entries in each column of Matrix B.
*
* @note This stage is needed to handle the offset of matrix product
* https://github.com/google/gemmlowp/blob/master/doc/low-precision.md
*/
class CpuGemmLowpMatrixBReductionKernel : public ICpuKernel<CpuGemmLowpMatrixBReductionKernel>
{
public:
/** Default constructor */
CpuGemmLowpMatrixBReductionKernel() = default;
ARM_COMPUTE_DISALLOW_COPY_ALLOW_MOVE(CpuGemmLowpMatrixBReductionKernel);
/** Initialise the kernel's input and output.
*
* @param[in] src Input tensor. Data type supported: Data type supported: QASYMM8/QASYMM8_SIGNED/QSYMM8/QSYMM8_PER_CHANNEL
* @param[out] dst Output row-vector of sums of all the entries in each column of mtx_b. Data type supported: S32
* @param[in] info Kernel metadata:
* - k (num_mtx_b_rows) Number of matrix B rows.
* - is_reshaped (is_transposed1xW) True if the input tensor is transposed 1xW.
* - scalar Scalar value to multiply each reduced row by.
* - mul_byscalar True if each reduced row must be multiplied by a scalar value.
*/
void configure(const ITensorInfo *src, ITensorInfo *dst, const GEMMLowpReductionKernelInfo &info);
/** Static function to check if given info will lead to a valid configuration
*
* Similar to CpuGemmLowpMatrixBReductionKernel::configure()
*
* @return a status
*/
static Status validate(const ITensorInfo *src, const ITensorInfo *dst, const GEMMLowpReductionKernelInfo &info);
// Inherited methods overridden:
void run_op(ITensorPack &tensors, const Window &window, const ThreadInfo &info) override;
const char *name() const override;
private:
/** Execution of the reduction kernel specialized on the input type
*
* @param[in] src Input tensor
* @param[in] dst Output tensor
* @param[in] window Execution window
* @param[in] info Thread-related information
*/
template <typename T>
void run_internal(const ITensor *src, ITensor *dst, const Window &window, const ThreadInfo &info);
/** Common signature for all reduction functions
*
* @param[in] src Input tensor
* @param[out] dst Output tensor
* @param[in] window Region on which to execute the kernel. (Must be a valid region of the window returned by window()).
*/
using CpuGemmLowpMatrixBReductionKernelPtr = void (CpuGemmLowpMatrixBReductionKernel::*)(const ITensor *src,
ITensor *dst,
const Window &window,
const ThreadInfo &info);
CpuGemmLowpMatrixBReductionKernelPtr _func{nullptr};
int32_t _k{0};
int32_t _scalar{0};
bool _mul_by_scalar{false};
};
} // namespace kernels
} // namespace cpu
} // namespace arm_compute
#endif /* ARM_COMPUTE_CPU_GEMMLOWP_REDUCTION_KERNEL_H */
|
