1 /* 2 * Copyright (c) 2019-2021 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_CL_GEMMLOWP_MATRIXMULTIPLY_RESHAPED_KERNEL_H 25 #define ARM_COMPUTE_CL_GEMMLOWP_MATRIXMULTIPLY_RESHAPED_KERNEL_H 26 27 #include "arm_compute/core/KernelDescriptors.h" 28 #include "src/core/common/Macros.h" 29 #include "src/gpu/cl/ClCompileContext.h" 30 #include "src/gpu/cl/IClKernel.h" 31 32 namespace arm_compute 33 { 34 namespace opencl 35 { 36 namespace kernels 37 { 38 /** OpenCL kernel to multiply matrices when both the input matrices LHS (src0) and RHS (src1) have been reshaped 39 * 40 * @note The input matrices @p src0 and @p src1 must be reshaped through: 41 * - @ref opencl::kernels::ClGemmReshapeLhsMatrixKernel 42 * - @ref opencl::kernels::ClGemmReshapeRhsMatrixKernel 43 */ 44 class ClGemmLowpMatrixMultiplyReshapedKernel : public IClKernel 45 { 46 public: 47 ClGemmLowpMatrixMultiplyReshapedKernel(); 48 ARM_COMPUTE_DISALLOW_COPY_ALLOW_MOVE(ClGemmLowpMatrixMultiplyReshapedKernel); 49 /** Initialise the kernel's input and dst. 50 * 51 * @param[in] compile_context The compile context to be used. 52 * @param[in] src0 Source tensor containing the LHS reshaped matrix. Data type supported: QASYMM8/QASYMM8_SIGNED. The number of dimensions for the LHS matrix must be less or equal than 4. 53 * @param[in] src1 Source tensor containing the RHS reshaped matrix. Data type supported: same as @p src0. The number of dimensions for the RHS matrix must be less or equal than 3. 54 * @param[out] dst Destination tensor to store the result of matrix multiplication. Data type supported: S32 55 * @param[in] lhs_info LHS matrix information used for reshaping the src0 tensor. Only the following values are supported: 56 * lhs_info.m0: 2,3,4,5,6,7,8 57 * lhs_info.k0: 2,3,4,8,16 58 * lhs_info.transpose: false 59 * @param[in] rhs_info RHS matrix information used for reshaping the src1 tensor. Only the following values are supported: 60 * rhs_info.n0: 2,3,4,8,16 61 * rhs_info.k0: same as lhs_info.k0 62 * rhs_info.transpose: true 63 * @param[in] gemm_info GEMM information used to retrieve the original dimensions of the input matrices 64 * 65 * @note lhs_info.k0 must be equal to rhs_info.k0 66 */ 67 void configure(const CLCompileContext &compile_context, const ITensorInfo *src0, const ITensorInfo *src1, ITensorInfo *dst, 68 const GEMMLHSMatrixInfo &lhs_info, const GEMMRHSMatrixInfo &rhs_info, const GEMMReshapeInfo &gemm_info); 69 /** Static function to check if given info will lead to a valid configuration 70 * 71 * Similar to @ref ClGemmLowpMatrixMultiplyReshapedKernel::configure() 72 * 73 * @return a status 74 */ 75 static Status validate(const ITensorInfo *src0, const ITensorInfo *src1, const ITensorInfo *dst, const GEMMLHSMatrixInfo &lhs_info, const GEMMRHSMatrixInfo &rhs_info, 76 const GEMMReshapeInfo &gemm_info); 77 78 // Inherited methods overridden: 79 void run_op(ITensorPack &tensors, const Window &window, cl::CommandQueue &queue) override; 80 81 private: 82 bool _slide_matrix_b{ true }; 83 bool _reinterpret_output_as_3d{ false }; 84 unsigned int _k{ 1 }; 85 bool _use_dummy_work_items{ false }; 86 }; 87 } // namespace kernels 88 } // namespace opencl 89 } // namespace arm_compute 90 #endif /* ARM_COMPUTE_CL_GEMMLOWP_MATRIXMULTIPLY_RESHAPED_KERNEL_H */ 91