1 /* 2 * Copyright (c) 2019-2022 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_CL_GEMM_HELPERS_H 25 #define ARM_COMPUTE_CL_GEMM_HELPERS_H 26 27 #include "arm_compute/core/TensorInfo.h" 28 #include "arm_compute/core/Types.h" 29 30 namespace arm_compute 31 { 32 namespace opencl 33 { 34 namespace kernels 35 { 36 namespace gemm 37 { 38 /** Configure @ref GEMMLHSMatrixInfo and @ref GEMMRHSMatrixInfo 39 * 40 * @param[in] m Number of rows (M) in the LHS matrix not reshaped 41 * @param[in] n Number of columns (N) in the RHS matrix not reshaped 42 * @param[in] m0 Number of rows processed by each thread/work-item 43 * @param[in] n0 Number of columns processed by each thread/work-item 44 * @param[in] k0 Number of inner accumulation performed by each thread/work-item 45 * @param[in] v0 Number of vertical blocks of size (m0xk0) stored on the same output row 46 * @param[in] h0 Number of horizontal blocks of size (k0xn0) stored on the same output row 47 * @param[in] lhs_interleave True if the v0 (m0xk0) blocks have to be interleaved in the output row 48 * @param[in] rhs_interleave True if the h0 (k0xn0) blocks have to be interleaved in the output row 49 * @param[in] lhs_transpose True if the (m0xk0) block has to be transposed before been stored 50 * @param[in] rhs_transpose True if the (k0xn0) block has to be transposed before been stored 51 * @param[in] export_to_cl_image (Optional) True if the RHS reshaped matrix has to be exported to cl_image 52 * 53 * @return @ref GEMMLHSMatrixInfo and @ref GEMMRHSMatrixInfo 54 */ 55 std::pair<GEMMLHSMatrixInfo, GEMMRHSMatrixInfo> configure_lhs_rhs_info(unsigned int m, unsigned int n, unsigned int m0, unsigned int n0, unsigned int k0, unsigned int v0, unsigned int h0, 56 bool lhs_interleave, bool rhs_interleave, bool lhs_transpose, bool rhs_transpose, bool export_to_cl_image = false); 57 58 /** Select @ref GEMMLHSMatrixInfo and @ref GEMMRHSMatrixInfo 59 * 60 * This function accepts two pairs of GEMMLHSMatrixInfo/GEMMRHSMatrixInfo where only the first is with cl_image2d support, 61 * and selects the valid one validating the GEMMRHSMatrixInfo. If the validation passes, the functions will return 62 * the first GEMMLHSMatrixInfo/GEMMRHSMatrixInfo pair with cl_image2d support. 63 * 64 * @param[in] info_img GEMMLHSMatrixInfo/GEMMRHSMatrixInfo with cl_image2d support 65 * @param[in] info_buf GEMMLHSMatrixInfo/GEMMRHSMatrixInfo to fall-back if cl_image2d cannot be used 66 * @param[in] n Number of columns (N) in the RHS matrix not reshaped 67 * @param[in] k Number of rows (K) in the RHS matrix not reshaped 68 * @param[in] b Batch size 69 * @param[in] data_type Data type 70 * 71 * @return @ref GEMMLHSMatrixInfo and @ref GEMMRHSMatrixInfo 72 */ 73 std::pair<GEMMLHSMatrixInfo, GEMMRHSMatrixInfo> select_lhs_rhs_info(std::pair<GEMMLHSMatrixInfo, GEMMRHSMatrixInfo> info_img, 74 std::pair<GEMMLHSMatrixInfo, GEMMRHSMatrixInfo> info_buf, 75 unsigned int n, unsigned int k, unsigned int b, DataType data_type); 76 77 /** Update padding required to export the OpenCL buffer to OpenCL image2d 78 * 79 * @param[in,out] tensor ITensorInfo of the tensor required to be exported to OpenCL image2d 80 */ 81 void update_padding_for_cl_image(ITensorInfo *tensor); 82 83 /** Utility function to validate the image2d OpenCL object support on the RHS reshaped matrix 84 * 85 * @param[in] tensor_reshaped_info TensorInfo for the RHS reshaped matrix 86 * @param[in] rhs_info @ref GEMMRHSMatrixInfo 87 * 88 * @return Status reporting if we can use the image2d OpenCL object on the RHS reshaped matrix 89 */ 90 Status validate_image2d_support_on_rhs(const ITensorInfo &tensor_reshaped_info, const GEMMRHSMatrixInfo &rhs_info); 91 92 /** Determine if the MMUL kernels should be preferred 93 * 94 * @param[in] m Number of rows of the LHS matrix 95 * @param[in] n Number of columns of the RHS matrix 96 * @param[in] k Number of columns of the LHS matrix, rows of the RHS matrix 97 * @param[in] b Batch size 98 * @param[in] data_type Data type FP32/FP16 99 * @param[in, out] best_m0 Suggested M0 (number of rows of the output block) for the kernel 100 * @param[in, out] best_n0 Suggested N0 (number of columns of the output block) for the kernel 101 * 102 * @return true if MMUL kernel is preferred over kernels w/o MMUL, false otherwise 103 */ 104 bool is_mmul_kernel_preferred(const unsigned int m, const unsigned int n, const unsigned int k, const unsigned int b, 105 const DataType data_type, unsigned int &best_m0, unsigned int &best_n0); 106 } // namespace gemm 107 } // namespace kernels 108 } // namespace opencl 109 } // namespace arm_compute 110 #endif /* ARM_COMPUTE_CL_GEMM_HELPERS_H */ 111