1 /* 2 * Copyright (c) 2017-2020 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_CLGEMMMATRIXVECTORMULTIPLYKERNEL_H 25 #define ARM_COMPUTE_CLGEMMMATRIXVECTORMULTIPLYKERNEL_H 26 27 #include "src/core/CL/ICLKernel.h" 28 29 namespace arm_compute 30 { 31 class ICLTensor; 32 33 /** Interface for the GEMM matrix vector multiply kernel. **/ 34 class CLGEMMMatrixVectorMultiplyKernel : public ICLKernel 35 { 36 public: 37 /** Default constructor */ 38 CLGEMMMatrixVectorMultiplyKernel(); 39 /** Prevent instances of this class from being copied (As this class contains pointers) */ 40 CLGEMMMatrixVectorMultiplyKernel(const CLGEMMMatrixVectorMultiplyKernel &) = delete; 41 /** Prevent instances of this class from being copied (As this class contains pointers) */ 42 CLGEMMMatrixVectorMultiplyKernel &operator=(const CLGEMMMatrixVectorMultiplyKernel &) = delete; 43 /** Allow instances of this class to be moved */ 44 CLGEMMMatrixVectorMultiplyKernel(CLGEMMMatrixVectorMultiplyKernel &&) = default; 45 /** Allow instances of this class to be moved */ 46 CLGEMMMatrixVectorMultiplyKernel &operator=(CLGEMMMatrixVectorMultiplyKernel &&) = default; 47 /** Set the input and output of the kernel. 48 * 49 * @param[in] input0 The reshaped input tensor. Data types supported: QASYMM8/QASYMM8_SIGNED/F16/F32 50 * @param[in] input1 The 2D reshaped weights tensor. Data type supported: Same as @p input. 51 * @param[out] output The output 2D tensor. Data types supported: Same as @p input, S32 for QASYMM8/QASYMM8_SIGNED. 52 */ 53 void configure(const ICLTensor *input0, const ICLTensor *input1, ICLTensor *output); 54 /** Set the input and output of the kernel. 55 * 56 * @param[in] compile_context The compile context to be used. 57 * @param[in] input0 The reshaped input tensor. Data types supported: QASYMM8/QASYMM8_SIGNED/F16/F32 58 * @param[in] input1 The 2D reshaped weights tensor. Data type supported: Same as @p input. 59 * @param[out] output The output 2D tensor. Data types supported: Same as @p input, S32 for QASYMM8/QASYMM8_SIGNED. 60 */ 61 void configure(const CLCompileContext &compile_context, const ICLTensor *input0, const ICLTensor *input1, ICLTensor *output); 62 /** Static function to check if given info will lead to a valid configuration of @ref CLGEMMMatrixVectorMultiplyKernel 63 * 64 * @param[in] input0 The reshaped input tensor info. Data types supported: QASYMM8/QASYMM8_SIGNED/F16/F32 65 * @param[in] input1 The 2D reshaped weights tensor info. Data type supported: Same as @p input. 66 * @param[in] output The output 2D tensor info. Data types supported: Same as @p input, S32 for QASYMM8/QASYMM8_SIGNED. 67 * 68 * @return a status 69 */ 70 static Status validate(const ITensorInfo *input0, const ITensorInfo *input1, const ITensorInfo *output); 71 72 // Inherited methods overridden: 73 void run(const Window &window, cl::CommandQueue &queue) override; 74 BorderSize border_size() const override; 75 76 private: 77 const ICLTensor *_input0; 78 const ICLTensor *_input1; 79 ICLTensor *_output; 80 int _num_rows_read_per_iteration; 81 BorderSize _border_size; 82 }; 83 } // arm_compute 84 #endif /*ARM_COMPUTE_CLGEMMMATRIXVECTORMULTIPLYKERNEL_H */ 85