From e75a02b60736f37c34388c23c0ccee230f65da59 Mon Sep 17 00:00:00 2001 From: Gian Marco Date: Wed, 8 Nov 2017 12:24:09 +0000 Subject: COMPMID-675 - Reworked NEGEMMLowp interface/function The new interface makes NEGEMMLowp able to work with ASYMM8 data types. Implemented 2 new functions: - NEGEMMLowpMatrixMultiplyCore - NEGEMMLowpOutputStage These functions should make the integration in android NN doable For more information about GEMMLowp: https://github.com/google/gemmlowp/blob/master/doc/low-precision.md Change-Id: Ie2c775f45234f68ca53dba644b3a912b997fd890 Reviewed-on: http://mpd-gerrit.cambridge.arm.com/95504 Tested-by: Kaizen Reviewed-by: Pablo Tello --- .../kernels/NEGEMMLowpOffsetContributionKernel.cpp | 338 +++++++++++++++++++++ 1 file changed, 338 insertions(+) create mode 100644 src/core/NEON/kernels/NEGEMMLowpOffsetContributionKernel.cpp (limited to 'src/core/NEON/kernels/NEGEMMLowpOffsetContributionKernel.cpp') diff --git a/src/core/NEON/kernels/NEGEMMLowpOffsetContributionKernel.cpp b/src/core/NEON/kernels/NEGEMMLowpOffsetContributionKernel.cpp new file mode 100644 index 0000000000..bd550db54c --- /dev/null +++ b/src/core/NEON/kernels/NEGEMMLowpOffsetContributionKernel.cpp @@ -0,0 +1,338 @@ +/* + * Copyright (c) 2017 ARM Limited. + * + * SPDX-License-Identifier: MIT + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to + * deal in the Software without restriction, including without limitation the + * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or + * sell copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in all + * copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, + * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + */ +#include "arm_compute/core/NEON/kernels/NEGEMMLowpOffsetContributionKernel.h" + +#include "arm_compute/core/AccessWindowStatic.h" +#include "arm_compute/core/Error.h" +#include "arm_compute/core/Helpers.h" +#include "arm_compute/core/ITensor.h" +#include "arm_compute/core/TensorInfo.h" +#include "arm_compute/core/Types.h" +#include "arm_compute/core/Utils.h" +#include "arm_compute/core/Validate.h" +#include "arm_compute/core/Window.h" + +#include +#include +#include + +using namespace arm_compute; + +namespace arm_compute +{ +class Coordinates; +} // namespace arm_compute + +NEGEMMLowpOffsetContributionKernel::NEGEMMLowpOffsetContributionKernel() + : _vector_sum_col(nullptr), _vector_sum_row(nullptr), _mm_result(nullptr), _a_offset(0), _b_offset(0), _k_offset(0), _slide_vector_sum_col(true) +{ +} + +void NEGEMMLowpOffsetContributionKernel::configure(ITensor *mm_result, const ITensor *vector_sum_col, const ITensor *vector_sum_row, int32_t k, int32_t a_offset, int32_t b_offset) +{ + ARM_COMPUTE_ERROR_ON_DATA_TYPE_CHANNEL_NOT_IN(mm_result, 1, DataType::S32); + + // If a_offset == 0, vector_sum_col can be a nullptr + if(a_offset != 0) + { + ARM_COMPUTE_ERROR_ON_DATA_TYPE_CHANNEL_NOT_IN(vector_sum_col, 1, DataType::S32); + ARM_COMPUTE_ERROR_ON(vector_sum_col->info()->dimension(0) != mm_result->info()->dimension(0)); + + TensorShape vector_sum_col_shape = vector_sum_col->info()->tensor_shape(); + vector_sum_col_shape.collapse(1); + + // Check if vector_sum_col_shape should be slidden or not + // Don't slide vector_sum_col_shape along the y dimension if vector_sum_col_shape has just 1 dimension and vector_sum_row_shape more than 1 + // This scenario can happen when the the matrix multiplication is used to perform a convolution operation + _slide_vector_sum_col = vector_sum_col_shape[1] != 1; + } + + // If b_offset == 0, vector_sum_row can be a nullptr + if(b_offset != 0) + { + ARM_COMPUTE_ERROR_ON_DATA_TYPE_CHANNEL_NOT_IN(vector_sum_row, 1, DataType::S32); + ARM_COMPUTE_ERROR_ON(vector_sum_row->info()->dimension(0) != mm_result->info()->dimension(1)); + + TensorShape output_shape = mm_result->info()->tensor_shape(); + TensorShape vector_sum_row_shape = vector_sum_row->info()->tensor_shape(); + vector_sum_row_shape.collapse(1); + output_shape.collapse(2); + + ARM_COMPUTE_ERROR_ON_MSG(vector_sum_row_shape[1] != output_shape[2], "mm_result tensor must have the same number of batches of output tensor"); + + if(a_offset != 0) + { + TensorShape vector_sum_col_shape = vector_sum_col->info()->tensor_shape(); + vector_sum_col_shape.collapse(1); + + ARM_COMPUTE_ERROR_ON_MSG(vector_sum_col_shape[1] != 1 + && vector_sum_col_shape[1] != vector_sum_row_shape[1], + "vector_sum_col tensor must have the same number of batches of vector_sum_row_shape or the number of batches must be set to 1"); + } + } + + _vector_sum_col = vector_sum_col; + _vector_sum_row = vector_sum_row; + _mm_result = mm_result; + _a_offset = a_offset; + _b_offset = b_offset; + _k_offset = a_offset * b_offset * k; + + constexpr unsigned int num_elems_processed_per_iteration = 16; + + // Configure kernel window + Window win = calculate_max_window(*mm_result->info(), Steps(num_elems_processed_per_iteration)); + + AccessWindowHorizontal mm_result_access(mm_result->info(), 0, num_elems_processed_per_iteration); + + // Accordingly with a_offset and b_offset, we can have 4 cases: + // a_offset != 0 && b_offset != 0 + // a_offset = 0 && b_offset != 0 + // a_offset != 0 && b_offset = 0 + // a_offset = 0 && b_offset = 0 + if(a_offset != 0 && b_offset != 0) + { + AccessWindowStatic vector_sum_row_access(vector_sum_row->info(), 0, 0, vector_sum_row->info()->dimension(0), 0); + AccessWindowHorizontal vector_sum_col_access(vector_sum_col->info(), 0, num_elems_processed_per_iteration); + + update_window_and_padding(win, + vector_sum_col_access, + vector_sum_row_access, + mm_result_access); + } + else if(a_offset == 0 && b_offset != 0) + { + AccessWindowStatic vector_sum_row_access(vector_sum_row->info(), 0, 0, vector_sum_row->info()->dimension(0), 0); + + update_window_and_padding(win, + vector_sum_row_access, + mm_result_access); + } + else if(a_offset != 0 && b_offset == 0) + { + AccessWindowHorizontal vector_sum_col_access(vector_sum_col->info(), 0, num_elems_processed_per_iteration); + + update_window_and_padding(win, + vector_sum_col_access, + mm_result_access); + } + else + { + update_window_and_padding(win, + mm_result_access); + } + + INEKernel::configure(win); +} + +void NEGEMMLowpOffsetContributionKernel::run(const Window &window, const ThreadInfo &info) +{ + ARM_COMPUTE_UNUSED(info); + ARM_COMPUTE_ERROR_ON_UNCONFIGURED_KERNEL(this); + ARM_COMPUTE_ERROR_ON_INVALID_SUBWINDOW(INEKernel::window(), window); + + Window collapsed_window = window.collapse_if_possible(IKernel::window(), Window::DimZ); + + if(_a_offset != 0 && _b_offset != 0) // true, true + { + // Set window for vector_sum_col + Window win_vector_sum_col(collapsed_window); + win_vector_sum_col.set(Window::DimY, Window::Dimension(0, 0, 0)); + if(!_slide_vector_sum_col) + { + win_vector_sum_col.set(Window::DimZ, Window::Dimension(0, 0, 0)); + } + + // Set window for vector_sum_row + Window win_vector_sum_row(collapsed_window); + win_vector_sum_row.set(Window::DimX, Window::Dimension(0, 0, 0)); + win_vector_sum_row.set(Window::DimY, Window::Dimension(0, 0, 0)); + + Iterator vector_sum_col(_vector_sum_col, win_vector_sum_col); + Iterator vector_sum_row(_vector_sum_row, win_vector_sum_row); + Iterator mm_result(_mm_result, window); + + execute_window_loop(window, [&](const Coordinates & id) + { + // Compute the leftover term due to a_offset. + int32x4x4_t a_offset_term_s32 = + { + { + vld1q_s32(reinterpret_cast(vector_sum_col.ptr()) + 0), + vld1q_s32(reinterpret_cast(vector_sum_col.ptr()) + 4), + vld1q_s32(reinterpret_cast(vector_sum_col.ptr()) + 8), + vld1q_s32(reinterpret_cast(vector_sum_col.ptr()) + 12) + } + }; + + a_offset_term_s32.val[0] = vmulq_n_s32(a_offset_term_s32.val[0], _a_offset); + a_offset_term_s32.val[1] = vmulq_n_s32(a_offset_term_s32.val[1], _a_offset); + a_offset_term_s32.val[2] = vmulq_n_s32(a_offset_term_s32.val[2], _a_offset); + a_offset_term_s32.val[3] = vmulq_n_s32(a_offset_term_s32.val[3], _a_offset); + + // Compute the leftover term due to b_offset. + int32x4_t b_offset_term_s32 = vld1q_dup_s32(reinterpret_cast(vector_sum_row.ptr()) + id.y()); + b_offset_term_s32 = vmulq_n_s32(b_offset_term_s32, _b_offset); + + // Add a_offset_term_s32 and b_offset_term_s32 + int32x4x4_t offset_term_s32 = + { + { + vdupq_n_s32(_k_offset), + vdupq_n_s32(_k_offset), + vdupq_n_s32(_k_offset), + vdupq_n_s32(_k_offset) + } + }; + + offset_term_s32.val[0] = vaddq_s32(offset_term_s32.val[0], vaddq_s32(a_offset_term_s32.val[0], b_offset_term_s32)); + offset_term_s32.val[1] = vaddq_s32(offset_term_s32.val[1], vaddq_s32(a_offset_term_s32.val[1], b_offset_term_s32)); + offset_term_s32.val[2] = vaddq_s32(offset_term_s32.val[2], vaddq_s32(a_offset_term_s32.val[2], b_offset_term_s32)); + offset_term_s32.val[3] = vaddq_s32(offset_term_s32.val[3], vaddq_s32(a_offset_term_s32.val[3], b_offset_term_s32)); + + int32x4x4_t in_s32 = + { + { + vld1q_s32(reinterpret_cast(mm_result.ptr()) + 0), + vld1q_s32(reinterpret_cast(mm_result.ptr()) + 4), + vld1q_s32(reinterpret_cast(mm_result.ptr()) + 8), + vld1q_s32(reinterpret_cast(mm_result.ptr()) + 12) + } + }; + + // Add the offset terms to GEMM's result + in_s32.val[0] = vaddq_s32(in_s32.val[0], offset_term_s32.val[0]); + in_s32.val[1] = vaddq_s32(in_s32.val[1], offset_term_s32.val[1]); + in_s32.val[2] = vaddq_s32(in_s32.val[2], offset_term_s32.val[2]); + in_s32.val[3] = vaddq_s32(in_s32.val[3], offset_term_s32.val[3]); + + // Store the result with the offset contribution + vst1q_s32(reinterpret_cast(mm_result.ptr()) + 0, in_s32.val[0]); + vst1q_s32(reinterpret_cast(mm_result.ptr()) + 4, in_s32.val[1]); + vst1q_s32(reinterpret_cast(mm_result.ptr()) + 8, in_s32.val[2]); + vst1q_s32(reinterpret_cast(mm_result.ptr()) + 12, in_s32.val[3]); + }, + vector_sum_col, vector_sum_row, mm_result); + } + else if((_a_offset == 0) && (_b_offset != 0)) // false, true + { + // Set window for vector_sum_row + Window win_vector_sum_row(collapsed_window); + win_vector_sum_row.set(Window::DimX, Window::Dimension(0, 0, 0)); + win_vector_sum_row.set(Window::DimY, Window::Dimension(0, 0, 0)); + + Iterator vector_sum_row(_vector_sum_row, win_vector_sum_row); + Iterator mm_result(_mm_result, window); + + execute_window_loop(window, [&](const Coordinates & id) + { + // Compute the leftover term due to b_offset. + int32x4_t b_offset_term_s32 = vld1q_dup_s32(reinterpret_cast(vector_sum_row.ptr()) + id.y()); + b_offset_term_s32 = vmulq_n_s32(b_offset_term_s32, _b_offset); + + int32x4x4_t in_s32 = + { + { + vld1q_s32(reinterpret_cast(mm_result.ptr()) + 0), + vld1q_s32(reinterpret_cast(mm_result.ptr()) + 4), + vld1q_s32(reinterpret_cast(mm_result.ptr()) + 8), + vld1q_s32(reinterpret_cast(mm_result.ptr()) + 12) + } + }; + + // Add the offset terms to GEMM's result + in_s32.val[0] = vaddq_s32(in_s32.val[0], b_offset_term_s32); + in_s32.val[1] = vaddq_s32(in_s32.val[1], b_offset_term_s32); + in_s32.val[2] = vaddq_s32(in_s32.val[2], b_offset_term_s32); + in_s32.val[3] = vaddq_s32(in_s32.val[3], b_offset_term_s32); + + // Store the result with the offset contribution + vst1q_s32(reinterpret_cast(mm_result.ptr()) + 0, in_s32.val[0]); + vst1q_s32(reinterpret_cast(mm_result.ptr()) + 4, in_s32.val[1]); + vst1q_s32(reinterpret_cast(mm_result.ptr()) + 8, in_s32.val[2]); + vst1q_s32(reinterpret_cast(mm_result.ptr()) + 12, in_s32.val[3]); + }, + vector_sum_row, mm_result); + } + else if((_a_offset != 0) && (_b_offset == 0)) // true, false + { + // Set window for vector_sum_col + Window win_vector_sum_col(collapsed_window); + win_vector_sum_col.set(Window::DimY, Window::Dimension(0, 0, 0)); + if(!_slide_vector_sum_col) + { + win_vector_sum_col.set(Window::DimZ, Window::Dimension(0, 0, 0)); + } + + Iterator vector_sum_col(_vector_sum_col, win_vector_sum_col); + Iterator mm_result(_mm_result, window); + + execute_window_loop(window, [&](const Coordinates & id) + { + // Compute the leftover term due to a_offset. + int32x4x4_t a_offset_term_s32 = + { + { + vld1q_s32(reinterpret_cast(vector_sum_col.ptr()) + 0), + vld1q_s32(reinterpret_cast(vector_sum_col.ptr()) + 4), + vld1q_s32(reinterpret_cast(vector_sum_col.ptr()) + 8), + vld1q_s32(reinterpret_cast(vector_sum_col.ptr()) + 12) + } + }; + + a_offset_term_s32.val[0] = vmulq_n_s32(a_offset_term_s32.val[0], _a_offset); + a_offset_term_s32.val[1] = vmulq_n_s32(a_offset_term_s32.val[1], _a_offset); + a_offset_term_s32.val[2] = vmulq_n_s32(a_offset_term_s32.val[2], _a_offset); + a_offset_term_s32.val[3] = vmulq_n_s32(a_offset_term_s32.val[3], _a_offset); + + int32x4x4_t in_s32 = + { + { + vld1q_s32(reinterpret_cast(mm_result.ptr()) + 0), + vld1q_s32(reinterpret_cast(mm_result.ptr()) + 4), + vld1q_s32(reinterpret_cast(mm_result.ptr()) + 8), + vld1q_s32(reinterpret_cast(mm_result.ptr()) + 12) + } + }; + + // Add the offset terms to GEMM's result + in_s32.val[0] = vaddq_s32(in_s32.val[0], a_offset_term_s32.val[0]); + in_s32.val[1] = vaddq_s32(in_s32.val[1], a_offset_term_s32.val[1]); + in_s32.val[2] = vaddq_s32(in_s32.val[2], a_offset_term_s32.val[2]); + in_s32.val[3] = vaddq_s32(in_s32.val[3], a_offset_term_s32.val[3]); + + // Store the result with the offset contribution + vst1q_s32(reinterpret_cast(mm_result.ptr()) + 0, in_s32.val[0]); + vst1q_s32(reinterpret_cast(mm_result.ptr()) + 4, in_s32.val[1]); + vst1q_s32(reinterpret_cast(mm_result.ptr()) + 8, in_s32.val[2]); + vst1q_s32(reinterpret_cast(mm_result.ptr()) + 12, in_s32.val[3]); + }, + vector_sum_col, mm_result); + } + else // false, false + { + // No offset contribution from matrix A and matrix B + return; + } +} -- cgit v1.2.1