1 /* 2 * Copyright (c) 2016-2020 Arm Limited. 3 * 4 * SPDX-License-Identifier: MIT 5 * 6 * Permission is hereby granted, free of charge, to any person obtaining a copy 7 * of this software and associated documentation files (the "Software"), to 8 * deal in the Software without restriction, including without limitation the 9 * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or 10 * sell copies of the Software, and to permit persons to whom the Software is 11 * furnished to do so, subject to the following conditions: 12 * 13 * The above copyright notice and this permission notice shall be included in all 14 * copies or substantial portions of the Software. 15 * 16 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR 17 * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, 18 * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE 19 * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER 20 * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, 21 * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE 22 * SOFTWARE. 23 */ 24 #ifndef ARM_COMPUTE_NEARITHMETICADDITIONKERNEL_H 25 #define ARM_COMPUTE_NEARITHMETICADDITIONKERNEL_H 26 27 #include "arm_compute/core/Types.h" 28 #include "src/core/NEON/INEKernel.h" 29 30 namespace arm_compute 31 { 32 class ITensor; 33 34 /** Interface for the kernel to perform addition between two tensors */ 35 class NEArithmeticAdditionKernel : public INEKernel 36 { 37 public: name()38 const char *name() const override 39 { 40 return "NEArithmeticAdditionKernel"; 41 } 42 /** Default constructor */ 43 NEArithmeticAdditionKernel(); 44 /** Prevent instances of this class from being copied (As this class contains pointers) */ 45 NEArithmeticAdditionKernel(const NEArithmeticAdditionKernel &) = delete; 46 /** Prevent instances of this class from being copied (As this class contains pointers) */ 47 NEArithmeticAdditionKernel &operator=(const NEArithmeticAdditionKernel &) = delete; 48 /** Allow instances of this class to be moved */ 49 NEArithmeticAdditionKernel(NEArithmeticAdditionKernel &&) = default; 50 /** Allow instances of this class to be moved */ 51 NEArithmeticAdditionKernel &operator=(NEArithmeticAdditionKernel &&) = default; 52 /** Default destructor */ 53 ~NEArithmeticAdditionKernel() = default; 54 55 /** Initialise the kernel's input, output and border mode. 56 * 57 * Valid configurations (Input1,Input2) -> Output : 58 * 59 * - (U8,U8) -> U8 60 * - (U8,U8) -> S16 61 * - (S16,U8) -> S16 62 * - (U8,S16) -> S16 63 * - (S16,S16) -> S16 64 * - (S32,S32) -> S32 65 * - (F16,F16) -> F16 66 * - (F32,F32) -> F32 67 * - (QASYMM8,QASYMM8) -> QASYMM8 68 * - (QASYMM8_SIGNED,QASYMM8_SIGNED) -> QASYMM8_SIGNED 69 * - (QSYMM16,QSYMM16) -> QSYMM16 70 * 71 * @param[in] input1 First input tensor info. Data types supported: U8/QASYMM8/QASYMM8_SIGNED/S16/QSYMM16/F16/S32/F32 72 * @param[in] input2 Second input tensor info. Data types supported: U8/QASYMM8/QASYMM8_SIGNED/S16/QSYMM16/F16/S32/F32 73 * @param[out] output The output tensor info. Data types supported: U8/QASYMM8/QASYMM8_SIGNED/S16/QSYMM16/F16/S32/F32. 74 * @param[in] policy Overflow policy. 75 */ 76 void configure(const ITensorInfo *input1, const ITensorInfo *input2, ITensorInfo *output, ConvertPolicy policy); 77 /** Static function to check if given info will lead to a valid configuration of @ref NEArithmeticAdditionKernel 78 * 79 * @param[in] input1 First input tensor info. Data types supported: U8/QASYMM8/QASYMM8_SIGNED/S16/QSYMM16/F16/S32/F32 80 * @param[in] input2 Second input tensor info. Data types supported: U8/QASYMM8/QASYMM8_SIGNED/S16/QSYMM16/F16/S32/F32 81 * @param[in] output The output tensor info. Data types supported: U8/QASYMM8/QASYMM8_SIGNED/S16/QSYMM16/F16/S32/F32. 82 * @param[in] policy Overflow policy. 83 * 84 * @return a status 85 */ 86 static Status validate(const ITensorInfo *input1, const ITensorInfo *input2, const ITensorInfo *output, ConvertPolicy policy); 87 88 // Inherited methods overridden: 89 void run_op(ITensorPack &tensors, const Window &window, const ThreadInfo &info) override; 90 91 private: 92 /** Common signature for all the specialised add functions 93 * 94 * @param[in] input1 First input tensor. Data types supported: U8/QASYMM8/S16/QSYMM16/F16/S32/F32 95 * @param[in] input2 Second input tensor. Data types supported: U8/QASYMM8/S16/QSYMM16/F16/S32/F32 96 * @param[out] output The output tensor. Data types supported: U8/QASYMM8/S16/QSYMM16/F16/S32/F32. 97 * @param[in] policy Overflow policy. 98 * @param[in] window Region on which to execute the kernel. 99 */ 100 using AddFunction = void(const ITensor *input1, const ITensor *input2, ITensor *output, ConvertPolicy policy, const Window &window); 101 /** Add function to use for the particular tensor types passed to configure() */ 102 AddFunction *_func; 103 ConvertPolicy _policy; 104 }; 105 } // namespace arm_compute 106 #endif /*ARM_COMPUTE_NEARITHMETICADDITIONKERNEL_H */ 107