aboutsummaryrefslogtreecommitdiff
path: root/arm_compute/core/NEON/kernels/NEElementwiseOperationKernel.h
blob: 93ad4373222c9a98924280c22d6f9cd2f7785004 (plain)
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
/*
 * Copyright (c) 2018 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, INNEUDING 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 NEAIM, 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_NEELEMENTWISEOPERATIONKERNEL_H__
#define __ARM_COMPUTE_NEELEMENTWISEOPERATIONKERNEL_H__

#include "arm_compute/core/NEON/INEKernel.h"
#include "arm_compute/core/Types.h"

namespace arm_compute
{
class ITensor;

/** Interface for an element-wise operation kernel
 *
 * Element-wise operation is computed by:
 * @f[ output(x,y) = OP(input1(x,y), input2(x,y))@f]
 *
 */
class NEElementwiseOperationKernel : public INEKernel
{
public:
    const char *name() const override
    {
        return "NEElementwiseOperationKernel";
    }
    /** Default constructor */
    NEElementwiseOperationKernel();
    /** Prevent instances of this class from being copied (As this class contains pointers) */
    NEElementwiseOperationKernel(const NEElementwiseOperationKernel &) = delete;
    /** Prevent instances of this class from being copied (As this class contains pointers) */
    NEElementwiseOperationKernel &operator=(const NEElementwiseOperationKernel &) = delete;
    /** Allow instances of this class to be moved */
    NEElementwiseOperationKernel(NEElementwiseOperationKernel &&) = default;
    /** Allow instances of this class to be moved */
    NEElementwiseOperationKernel &operator=(NEElementwiseOperationKernel &&) = default;
    /** Default destructor */
    ~NEElementwiseOperationKernel() = default;

    // Inherited methods overridden:
    void run(const Window &window, const ThreadInfo &info) override;

protected:
    /** Validate the argument passed to the kernel
     *
     * @param[in] input1 First tensor input. Data types supported: S16/F16/S32/F32.
     * @param[in] input2 Second tensor input. Data types supported: Same as @p input1.
     * @param[in] output Output tensor. Data types supported: Same as @p input1.
     */
    virtual Status validate_arguments(const ITensorInfo &input1, const ITensorInfo &input2, const ITensorInfo &output) = 0;

    /** Commmon configure function for element-wise operators with no additional options (e.g. Min, Max, SquaredDiff)
     *
     */
    template <ArithmeticOperation op>
    void configure_common(const ITensor *input1, const ITensor *input2, ITensor *output);

    ArithmeticOperation _op; // Code of the operation to execute

private:
    /** Common signature for all the specialised add functions
     *
     * @param[in]  input1 An input tensor. Data types supported: S16/F16/S32/F32
     * @param[in]  input2 An input tensor. Data types supported: S16/F16/S32/F32
     * @param[out] output The output tensor. Data types supported: S16/F16/S32/F32
     * @param[in]  window Region on which to execute the kernel.
     */
    using ElementwiseFunction = void(const ITensor *input1, const ITensor *input2, ITensor *output, const Window &window);
    /** Add function to use for the particular tensor types passed to configure() */
    ElementwiseFunction *_func;
    const ITensor       *_input1;
    const ITensor       *_input2;
    ITensor             *_output;
};

class NEArithmeticOperationKernel : public NEElementwiseOperationKernel
{
public:
    NEArithmeticOperationKernel()
        : NEElementwiseOperationKernel()
    {
    }

    /** Static function to check if given info will lead to a valid configuration of @ref NEArithmeticOperationKernel
     *
     * @param[in] op     Arithmetic operation to be executed.
     * @param[in] input1 First tensor input. Data types supported: QASYMM8/S16/F16/S32/F32.
     * @param[in] input1 First tensor input. Data types supported: Same as @p input1.
     * @param[in] input2 Second tensor input. Data types supported: Same as @p input1.
     * @param[in] output Output tensor. Data types supported: Same as @p input1.
     */
    void configure(ArithmeticOperation op, const ITensor *input1, const ITensor *input2, ITensor *output);

    /** Static function to check if given info will lead to a valid configuration of @ref NEArithmeticOperationKernel
     *
     * @param[in] op     Arithmetic operation to be executed.
     * @param[in] input1 First tensor input info. Data types supported: QASYMM8/S16/F16/S32/F32.
     * @param[in] input1 First tensor input info. Data types supported: Same as @p input1
     * @param[in] input2 Second tensor input info. Data types supported: Same as @p input1.
     * @param[in] output Output tensor info. Data types supported: Same as @p input1.
     *
     * @return a Status
     */
    static Status validate(ArithmeticOperation op, const ITensorInfo *input1, const ITensorInfo *input2, const ITensorInfo *output);

protected:
    // Inherited methods overridden:
    Status validate_arguments(const ITensorInfo &input1, const ITensorInfo &input2, const ITensorInfo &output) override;
};
} // namespace arm_compute
#endif /* __ARM_COMPUTE_NEELEMENTWISEOPERATIONKERNEL_H__ */