1 /* 2 * Copyright (c) 2016-2022 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_CPU_ADD_KERNEL_H 25 #define ARM_COMPUTE_CPU_ADD_KERNEL_H 26 27 #include "src/core/common/Macros.h" 28 #include "src/cpu/ICpuKernel.h" 29 30 namespace arm_compute 31 { 32 namespace cpu 33 { 34 namespace kernels 35 { 36 /** Interface for the kernel to perform addition between two tensors */ 37 class CpuAddKernel : public ICpuKernel<CpuAddKernel> 38 { 39 private: 40 using AddKernelPtr = std::add_pointer<void(const ITensor *, const ITensor *, ITensor *, const ConvertPolicy &, const Window &)>::type; 41 42 public: 43 struct AddKernel 44 { 45 const char *name; 46 const CpuAddKernelDataTypeISASelectorDataPtr is_selected; 47 AddKernelPtr ukernel; 48 }; 49 50 CpuAddKernel() = default; 51 ARM_COMPUTE_DISALLOW_COPY_ALLOW_MOVE(CpuAddKernel); 52 /** Initialise the kernel's input, dst and border mode. 53 * 54 * Valid configurations (src0,src1) -> dst : 55 * 56 * - (U8,U8) -> U8 57 * - (S16,S16) -> S16 58 * - (S32,S32) -> S32 59 * - (F16,F16) -> F16 60 * - (F32,F32) -> F32 61 * - (QASYMM8,QASYMM8) -> QASYMM8 62 * - (QASYMM8_SIGNED,QASYMM8_SIGNED) -> QASYMM8_SIGNED 63 * - (QSYMM16,QSYMM16) -> QSYMM16 64 * 65 * @param[in] src0 First input tensor info. Data types supported: U8/QASYMM8/QASYMM8_SIGNED/S16/QSYMM16/F16/S32/F32 66 * @param[in] src1 Second input tensor info. Data types supported: U8/QASYMM8/QASYMM8_SIGNED/S16/QSYMM16/F16/S32/F32 67 * @param[out] dst The dst tensor info. Data types supported: U8/QASYMM8/QASYMM8_SIGNED/S16/QSYMM16/F16/S32/F32. 68 * @param[in] policy Overflow policy. 69 */ 70 void configure(const ITensorInfo *src0, const ITensorInfo *src1, ITensorInfo *dst, ConvertPolicy policy); 71 /** Static function to check if given info will lead to a valid configuration 72 * 73 * Similar to CpuAddKernel::configure() 74 * 75 * @return a status 76 */ 77 static Status validate(const ITensorInfo *src0, const ITensorInfo *src1, const ITensorInfo *dst, ConvertPolicy policy); 78 79 // Inherited methods overridden: 80 void run_op(ITensorPack &tensors, const Window &window, const ThreadInfo &info) override; 81 const char *name() const override; 82 83 /** Return minimum workload size of the relevant kernel 84 * 85 * @param[in] platform The CPU platform used to create the context. 86 * @param[in] thread_count Number of threads in the execution. 87 * 88 * @return[out] mws Minimum workload size for requested configuration. 89 */ 90 size_t get_mws(const CPUInfo &platform, size_t thread_count) const override; 91 92 static const std::vector<AddKernel> &get_available_kernels(); 93 get_split_dimension()94 size_t get_split_dimension() const 95 { 96 return _split_dimension; 97 } 98 99 private: 100 ConvertPolicy _policy{}; 101 AddKernelPtr _run_method{ nullptr }; 102 std::string _name{}; 103 size_t _split_dimension{ Window::DimY }; 104 }; 105 } // namespace kernels 106 } // namespace cpu 107 } // namespace arm_compute 108 #endif /* ARM_COMPUTE_CPU_ADD_KERNEL_H */ 109