/* * Copyright (c) 2018-2019 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/CL/kernels/CLDepthwiseConvolutionLayer3x3NHWCKernel.h" #include "arm_compute/core/AccessWindowStatic.h" #include "arm_compute/core/CL/CLHelpers.h" #include "arm_compute/core/CL/CLKernelLibrary.h" #include "arm_compute/core/CL/CLValidate.h" #include "arm_compute/core/CL/ICLKernel.h" #include "arm_compute/core/CL/ICLTensor.h" #include "arm_compute/core/Error.h" #include "arm_compute/core/Helpers.h" #include "arm_compute/core/TensorInfo.h" #include "arm_compute/core/Types.h" #include "arm_compute/core/Utils.h" #include "arm_compute/core/utils/misc/ShapeCalculator.h" #include "arm_compute/core/utils/quantization/AsymmHelpers.h" namespace arm_compute { namespace { Status validate_arguments(const ITensorInfo *input, const ITensorInfo *weights, const ITensorInfo *biases, const ITensorInfo *output, const PadStrideInfo &conv_info, unsigned int depth_multiplier, const ActivationLayerInfo &act_info, const Size2D &dilation) { ARM_COMPUTE_RETURN_ERROR_ON_F16_UNSUPPORTED(input); ARM_COMPUTE_RETURN_ERROR_ON_DATA_TYPE_CHANNEL_NOT_IN(input, 1, DataType::F16, DataType::F32, DataType::QASYMM8); ARM_COMPUTE_RETURN_ERROR_ON_MSG((act_info.enabled()) && (input->data_type() == DataType::QASYMM8) && (act_info.activation() != ActivationLayerInfo::ActivationFunction::LU_BOUNDED_RELU) && (act_info.activation() != ActivationLayerInfo::ActivationFunction::BOUNDED_RELU) && (act_info.activation() != ActivationLayerInfo::ActivationFunction::RELU) && (act_info.activation() != ActivationLayerInfo::ActivationFunction::LOGISTIC), "For QASYMM8 only logistic, relu, lower bounded relu and lower-upper bounded relu are supported"); ARM_COMPUTE_RETURN_ERROR_ON_MISMATCHING_DATA_TYPES(input, weights); ARM_COMPUTE_RETURN_ERROR_ON(depth_multiplier > 1); // COMPMID-1071 Add depth multiplier support for NHWC ARM_COMPUTE_RETURN_ERROR_ON(conv_info.stride().first < 1); ARM_COMPUTE_RETURN_ERROR_ON(std::max(conv_info.pad_top(), conv_info.pad_bottom()) > 4); ARM_COMPUTE_RETURN_ERROR_ON((dilation.x() < 1) || (dilation.y() < 1)); const bool is_qasymm = is_data_type_quantized_asymmetric(input->data_type()); const size_t weights_width = 3; const size_t weights_height = 3; if(is_qasymm) { DepthwiseConvolutionReshapeInfo info; info.c0 = 4; ARM_COMPUTE_RETURN_ERROR_ON((weights->dimension(0) / info.c0) != weights_width * weights_height); } else { ARM_COMPUTE_RETURN_ERROR_ON((weights->dimension(1) != weights_width) || (weights->dimension(2) != weights_height)); } if(biases != nullptr) { if(is_qasymm) { ARM_COMPUTE_RETURN_ERROR_ON_DATA_TYPE_CHANNEL_NOT_IN(biases, 1, DataType::S32); } else { ARM_COMPUTE_RETURN_ERROR_ON(biases->dimension(0) != weights->dimension(0)); ARM_COMPUTE_RETURN_ERROR_ON_MISMATCHING_DATA_TYPES(weights, biases); } ARM_COMPUTE_RETURN_ERROR_ON(biases->num_dimensions() > 1); } if(output->total_size() != 0) { const TensorShape output_shape = arm_compute::misc::shape_calculator::compute_depthwise_convolution_shape( *input, TensorInfo(TensorShape(weights_width, weights_height), 1, weights->data_type()).set_data_layout(DataLayout::NCHW), conv_info, depth_multiplier, dilation); ARM_COMPUTE_RETURN_ERROR_ON_MISMATCHING_DIMENSIONS(output->tensor_shape(), output_shape); } if(is_qasymm) { const UniformQuantizationInfo iq_info = input->quantization_info().uniform(); const UniformQuantizationInfo wq_info = weights->quantization_info().uniform(); const UniformQuantizationInfo oq_info = (output->total_size() != 0) ? output->quantization_info().uniform() : iq_info; float multiplier = iq_info.scale * wq_info.scale / oq_info.scale; ARM_COMPUTE_UNUSED(multiplier); ARM_COMPUTE_RETURN_ERROR_ON(multiplier > 1.0f); } return Status{}; } std::pair validate_and_configure_window(ITensorInfo *input, ITensorInfo *weights, ITensorInfo *bias, ITensorInfo *output, const PadStrideInfo &conv_info, unsigned int depth_multiplier, const Size2D &dilation) { const size_t weights_width = 3; const size_t weights_height = 3; // Get convolved dimensions const TensorShape output_shape = arm_compute::misc::shape_calculator::compute_depthwise_convolution_shape( *input, TensorInfo(TensorShape(weights_width, weights_height), 1, weights->data_type()).set_data_layout(DataLayout::NCHW), conv_info, depth_multiplier, dilation); auto_init_if_empty(*output, input->clone()->set_tensor_shape(output_shape).set_quantization_info(output->quantization_info())); const bool is_qasymm = is_data_type_quantized_asymmetric(input->data_type()); const bool is_stride_1_dilation_1 = ((conv_info.stride().first == conv_info.stride().second) && (conv_info.stride().first == 1) && dilation.x() == 1 && dilation.y() == 1); const unsigned int num_rows_processed_per_iteration = is_stride_1_dilation_1 ? 2 : 1; const unsigned int num_elems_accessed_per_iteration = is_qasymm ? 4 : (8 / input->element_size()); const unsigned int num_rows_read_per_iteration = num_rows_processed_per_iteration + 2; const unsigned int num_rows_written_per_iteration = std::ceil(num_rows_processed_per_iteration / static_cast(conv_info.stride().first)); BorderSize border_size; border_size = BorderSize(conv_info.pad_left(), 0, std::max(std::max(conv_info.pad_right(), conv_info.pad_bottom()), conv_info.pad_top()), 0); // Configure kernel window Window win = calculate_max_window(*output, Steps(num_elems_accessed_per_iteration, num_rows_written_per_iteration)); AccessWindowStatic input_access(input, 0, -border_size.top, ceil_to_multiple(input->dimension(0), num_elems_accessed_per_iteration), ceil_to_multiple(input->dimension(1) + border_size.bottom, num_rows_read_per_iteration)); AccessWindowRectangle output_access(output, 0, 0, num_elems_accessed_per_iteration, num_rows_written_per_iteration); bool window_changed = false; if(is_qasymm) { window_changed = update_window_and_padding(win, input_access, output_access); } else { AccessWindowStatic weights_access(weights, 0, 0, ceil_to_multiple(weights->dimension(0), num_elems_accessed_per_iteration), weights->dimension(1)); window_changed = update_window_and_padding(win, input_access, weights_access, output_access); } if(bias != nullptr) { AccessWindowHorizontal bias_access(bias, 0, num_elems_accessed_per_iteration); window_changed = window_changed || update_window_and_padding(win, bias_access); } output_access.set_valid_region(win, ValidRegion(Coordinates(), output->tensor_shape())); Status err = (window_changed) ? ARM_COMPUTE_CREATE_ERROR(ErrorCode::RUNTIME_ERROR, "Insufficient Padding!") : Status{}; return std::make_pair(err, win); } } // namespace CLDepthwiseConvolutionLayer3x3NHWCKernel::CLDepthwiseConvolutionLayer3x3NHWCKernel() : _num_rows_processed_per_iteration(1), _num_planes_processed_per_iteration(1) { } BorderSize CLDepthwiseConvolutionLayer3x3NHWCKernel::border_size() const { return _border_size; } void CLDepthwiseConvolutionLayer3x3NHWCKernel::configure(const ICLTensor *input, const ICLTensor *weights, const ICLTensor *biases, ICLTensor *output, const PadStrideInfo &conv_info, unsigned int depth_multiplier, ActivationLayerInfo act_info, const Size2D &dilation) { ARM_COMPUTE_ERROR_ON_NULLPTR(input, weights, output); ARM_COMPUTE_ERROR_THROW_ON(validate_arguments(input->info(), weights->info(), (biases != nullptr) ? biases->info() : nullptr, output->info(), conv_info, depth_multiplier, act_info, dilation)); auto win_config = validate_and_configure_window(input->info(), weights->info(), biases != nullptr ? biases->info() : nullptr, output->info(), conv_info, depth_multiplier, dilation); ARM_COMPUTE_ERROR_THROW_ON(win_config.first); const bool is_qasymm = is_data_type_quantized_asymmetric(input->info()->data_type()); const bool is_stride_1 = ((conv_info.stride().first == conv_info.stride().second) && (conv_info.stride().first == 1)); const bool is_stride_1_dilation_1 = (is_stride_1 && dilation.x() == 1 && dilation.y() == 1); const bool is_dot8_supported = dot8_supported(CLKernelLibrary::get().get_device()); _input = input; _output = output; _weights = weights; _biases = biases; _conv_stride_y = conv_info.stride().second; _num_rows_processed_per_iteration = is_stride_1_dilation_1 ? 2 : 1; _num_planes_processed_per_iteration = is_stride_1_dilation_1 ? 2 : 1; // If QASYMM8 and the 8 bit dot product is available, force _num_planes_processed_per_iteration to 1 if(is_dot8_supported && is_qasymm) { _num_planes_processed_per_iteration = 1; } _border_size = BorderSize(is_qasymm && is_stride_1 ? 0 : conv_info.pad_left(), 0, std::max(std::max(conv_info.pad_right(), conv_info.pad_bottom()), conv_info.pad_top()), 0); const unsigned int num_elems_accessed_per_iteration = is_qasymm ? 4 : (8 / input->info()->element_size()); CLBuildOptions build_opts; build_opts.add_option("-DACTIVATION_TYPE=" + lower_string(string_from_activation_func(act_info.activation()))); build_opts.add_option_if(_biases != nullptr, "-DHAS_BIAS"); build_opts.add_option("-DVEC_SIZE=" + support::cpp11::to_string(num_elems_accessed_per_iteration)); build_opts.add_option("-DSRC_DIM_2=" + support::cpp11::to_string(_input->info()->dimension(2))); build_opts.add_option("-DCONV_PAD_TOP=" + support::cpp11::to_string(conv_info.pad_top())); build_opts.add_option("-DCONV_PAD_LEFT=" + support::cpp11::to_string(conv_info.pad_left())); build_opts.add_option("-DDILATION_X=" + support::cpp11::to_string(dilation.x())); build_opts.add_option("-DDILATION_Y=" + support::cpp11::to_string(dilation.y())); if(is_qasymm) { const UniformQuantizationInfo iq_info = _input->info()->quantization_info().uniform(); const UniformQuantizationInfo wq_info = _weights->info()->quantization_info().uniform(); const UniformQuantizationInfo oq_info = _output->info()->quantization_info().uniform(); float multiplier = iq_info.scale * wq_info.scale / oq_info.scale; int output_multiplier = 0; int output_shift = 0; quantization::calculate_quantized_multiplier_less_than_one(multiplier, &output_multiplier, &output_shift); build_opts.add_option("-DSRC_DIM_1=" + support::cpp11::to_string(_input->info()->dimension(1))); build_opts.add_option("-DINPUT_OFFSET=" + support::cpp11::to_string(-iq_info.offset)); build_opts.add_option("-DWEIGHTS_OFFSET=" + support::cpp11::to_string(-wq_info.offset)); build_opts.add_option("-DOUTPUT_OFFSET=" + support::cpp11::to_string(oq_info.offset)); build_opts.add_option("-DK_OFFSET=" + support::cpp11::to_string(9 * iq_info.offset * wq_info.offset)); build_opts.add_option("-DOUTPUT_MULTIPLIER=" + support::cpp11::to_string(output_multiplier)); build_opts.add_option("-DOUTPUT_SHIFT=" + support::cpp11::to_string(output_shift)); if(act_info.enabled()) { const int a_val = quantize_qasymm8(act_info.a(), oq_info); const int b_val = quantize_qasymm8(act_info.b(), oq_info); const int o1 = oq_info.offset; build_opts.add_option("-DA_VAL=" + support::cpp11::to_string(a_val)); build_opts.add_option("-DB_VAL=" + support::cpp11::to_string(b_val)); build_opts.add_option("-DCONST_0=" + support::cpp11::to_string(o1)); const float s1 = iq_info.scale; build_opts.add_option("-DS1_VAL=" + float_to_string_with_full_precision(s1)); build_opts.add_option("-DO1_VAL=" + support::cpp11::to_string(o1)); } } else { build_opts.add_option_if(act_info.enabled(), "-DA_VAL=" + float_to_string_with_full_precision(act_info.a())); build_opts.add_option_if(act_info.enabled(), "-DB_VAL=" + float_to_string_with_full_precision(act_info.b())); build_opts.add_option("-DDATA_TYPE=" + get_cl_type_from_data_type(_input->info()->data_type())); } if(is_stride_1_dilation_1) { build_opts.add_option("-DNUM_ROWS_PROCESSED=" + support::cpp11::to_string(_num_rows_processed_per_iteration)); build_opts.add_option("-DNUM_PLANES_PROCESSED=" + support::cpp11::to_string(_num_planes_processed_per_iteration)); build_opts.add_option("-DDST_DIM_2=" + support::cpp11::to_string(_output->info()->dimension(2))); } else { build_opts.add_option("-DCONV_STRIDE_X=" + support::cpp11::to_string(conv_info.stride().first)); build_opts.add_option("-DCONV_STRIDE_Y=" + support::cpp11::to_string(_conv_stride_y)); } build_opts.add_option_if(_input->info()->tensor_shape().total_size_upper(3) > 1, "-DDST_DEPTH=" + support::cpp11::to_string(static_cast(std::ceil(_output->info()->dimension(2) / static_cast(_num_planes_processed_per_iteration))))); std::string kernel_name; // Create kernel if(is_qasymm) { kernel_name = std::string("dwc_3x3_reshaped_qasymm8"); kernel_name += (is_dot8_supported && is_stride_1_dilation_1 ? "_dot8" : ""); kernel_name += (is_stride_1_dilation_1 ? "_stride1" : ""); kernel_name += "_nhwc"; } else { kernel_name = std::string("depthwise_convolution_3x3_nhwc"); kernel_name += (is_stride_1_dilation_1 ? "_stride1" : ""); } build_opts.add_option_if(input->info()->data_type() == DataType::F16, "-DIS_F16"); build_opts.add_option_if(input->info()->data_type() == DataType::F32, "-DIS_F32"); ICLKernel::configure_internal(win_config.second); _kernel = static_cast(CLKernelLibrary::get().create_kernel(kernel_name, build_opts.options())); // Set config_id for enabling LWS tuning _config_id = kernel_name; _config_id += "_"; _config_id += support::cpp11::to_string(input->info()->dimension(0)); _config_id += "_"; _config_id += support::cpp11::to_string(input->info()->dimension(1)); _config_id += "_"; _config_id += support::cpp11::to_string(input->info()->dimension(2)); _config_id += "_"; _config_id += support::cpp11::to_string(output->info()->dimension(0)); _config_id += "_"; _config_id += support::cpp11::to_string(output->info()->dimension(1)); _config_id += "_"; _config_id += string_from_data_type(input->info()->data_type()); } Status CLDepthwiseConvolutionLayer3x3NHWCKernel::validate(const ITensorInfo *input, const ITensorInfo *weights, const ITensorInfo *biases, const ITensorInfo *output, const PadStrideInfo &conv_info, unsigned int depth_multiplier, ActivationLayerInfo act_info, const Size2D &dilation) { ARM_COMPUTE_RETURN_ON_ERROR(validate_arguments(input, weights, biases, output, conv_info, depth_multiplier, act_info, dilation)); ARM_COMPUTE_RETURN_ON_ERROR(validate_and_configure_window(input->clone().get(), weights->clone().get(), biases != nullptr ? biases->clone().get() : nullptr, output->clone().get(), conv_info, depth_multiplier, dilation) .first); return Status{}; } void CLDepthwiseConvolutionLayer3x3NHWCKernel::run(const Window &window, cl::CommandQueue &queue) { ARM_COMPUTE_ERROR_ON_UNCONFIGURED_KERNEL(this); ARM_COMPUTE_ERROR_ON_INVALID_SUBWINDOW(IKernel::window(), window); // Collapse window Window window_collapsed = window.collapse_if_possible(ICLKernel::window(), Window::DimZ); const size_t total_batches = _input->info()->tensor_shape().total_size_upper(3); const bool is_qasymm = is_data_type_quantized_asymmetric(_input->info()->data_type()); Window win = window_collapsed; win.set(Window::DimZ, Window::Dimension(0, std::ceil(_output->info()->dimension(2) / static_cast(_num_planes_processed_per_iteration)) * total_batches, 1)); // Create input window and adjust Window win_in = win; win_in.set_dimension_step(Window::DimY, _num_rows_processed_per_iteration); win_in.set_dimension_step(Window::DimZ, _conv_stride_y); ARM_COMPUTE_ERROR_ON((win_in.y().step() < window.y().step()) || (win_in.z().step() < window.z().step())); Window slice_in = win_in.first_slice_window_4D(); Window slice_out = win.first_slice_window_4D(); unsigned int idx = 2 * num_arguments_per_4D_tensor() + (is_qasymm ? num_arguments_per_2D_tensor() : num_arguments_per_3D_tensor()); if(_biases != nullptr) { Window win_biases; win_biases.use_tensor_dimensions(_biases->info()->tensor_shape()); win_biases.set_dimension_step(Window::DimX, window.x().step()); add_1D_tensor_argument(idx, _biases, win_biases); } // Calculate the max_offset. // max_offset is the offset for the last NOT valid value in the Z dimension (spatial dimension Y for NHWC) // |******************| // | pad_top | // |******************| // | | // | plane0 | // | batch0 | // |__________________| // |******************| Batch 0 // | pad_bottom | // | pad_top | // |******************| // | | // | plane1 | // | batch0 | // |__________________|-----> max_offset // |******************| // | pad_bottom | // | pad_top | // |******************| // | | // | plane0 | // | batch1 | // |__________________| // |******************| Batch 1 // | pad_bottom | // | pad_top | // |******************| // | | // | plane1 | // | batch1 | // |__________________| // | pad_bottom | // |******************| const int max_offset = _input->info()->strides_in_bytes().z() * _input->info()->dimension(2) - (_input->info()->padding().bottom + _input->info()->padding().top) * _input->info()->strides_in_bytes().y(); _kernel.setArg(idx, max_offset); do { unsigned int idx = 0; add_4D_tensor_argument(idx, _input, slice_in); add_4D_tensor_argument(idx, _output, slice_out); if(is_qasymm) { add_2D_tensor_argument(idx, _weights, slice_out); } else { add_3D_tensor_argument(idx, _weights, slice_out); } enqueue(queue, *this, slice_out, lws_hint()); } while(win.slide_window_slice_4D(slice_out) && win_in.slide_window_slice_4D(slice_in)); } } // namespace arm_compute