/* * Copyright (c) 2018-2021 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 "src/core/CL/kernels/CLDepthwiseConvolutionLayer3x3NHWCKernel.h" #include "arm_compute/core/CL/CLHelpers.h" #include "arm_compute/core/CL/CLKernelLibrary.h" #include "arm_compute/core/CL/ICLTensor.h" #include "arm_compute/core/Helpers.h" #include "arm_compute/core/TensorInfo.h" #include "arm_compute/core/Utils.h" #include "arm_compute/core/utils/misc/ShapeCalculator.h" #include "arm_compute/core/utils/quantization/AsymmHelpers.h" #include "src/core/AccessWindowStatic.h" #include "src/core/CL/CLValidate.h" #include "src/core/CL/ICLKernel.h" #include "src/core/helpers/AutoConfiguration.h" #include "src/core/helpers/WindowHelpers.h" #include "support/StringSupport.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, const ITensorInfo *output_multipliers, const ITensorInfo *output_shifts) { 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, DataType::QASYMM8_SIGNED); ARM_COMPUTE_RETURN_ERROR_ON_MSG((act_info.enabled()) && (input->data_type() == DataType::QASYMM8 || input->data_type() == DataType::QASYMM8_SIGNED) && (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(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; const ConvolutionInfo info{ conv_info, depth_multiplier, ActivationLayerInfo(), dilation }; 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), info); if(is_qasymm) { DepthwiseConvolutionReshapeInfo info; info.c0 = 4; ARM_COMPUTE_RETURN_ERROR_ON((weights->dimension(0) / info.c0) != weights_width * weights_height); ARM_COMPUTE_RETURN_ERROR_ON_NULLPTR(output_multipliers, output_shifts); ARM_COMPUTE_RETURN_ERROR_ON_DATA_TYPE_CHANNEL_NOT_IN(output_multipliers, 1, DataType::S32); ARM_COMPUTE_RETURN_ERROR_ON_DATA_TYPE_CHANNEL_NOT_IN(output_shifts, 1, DataType::S32); ARM_COMPUTE_RETURN_ERROR_ON(output_multipliers->num_dimensions() > 1); ARM_COMPUTE_RETURN_ERROR_ON(output_shifts->num_dimensions() > 1); if(is_data_type_quantized_per_channel(weights->data_type())) { ARM_COMPUTE_RETURN_ERROR_ON(output_shape[0] != output_multipliers->dimension(0)); ARM_COMPUTE_RETURN_ERROR_ON(output_shape[0] != output_shifts->dimension(0)); } else { ARM_COMPUTE_RETURN_ERROR_ON_MISMATCHING_DATA_TYPES(input, weights); ARM_COMPUTE_RETURN_ERROR_ON(1 != output_multipliers->dimension(0)); ARM_COMPUTE_RETURN_ERROR_ON(1 != output_shifts->dimension(0)); } } else { ARM_COMPUTE_RETURN_ERROR_ON_MISMATCHING_DATA_TYPES(input, weights); ARM_COMPUTE_RETURN_ERROR_ON((weights->dimension(1) != weights_width) || (weights->dimension(2) != weights_height)); } if(biases != nullptr) { ARM_COMPUTE_RETURN_ERROR_ON(biases->dimension(0) != output_shape[0]); if(is_qasymm) { ARM_COMPUTE_RETURN_ERROR_ON_DATA_TYPE_CHANNEL_NOT_IN(biases, 1, DataType::S32); } else { ARM_COMPUTE_RETURN_ERROR_ON_MISMATCHING_DATA_TYPES(weights, biases); } ARM_COMPUTE_RETURN_ERROR_ON(biases->num_dimensions() > 1); } if(output->total_size() != 0) { ARM_COMPUTE_RETURN_ERROR_ON_MISMATCHING_DIMENSIONS(output->tensor_shape(), output_shape); } 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, ITensorInfo *output_multipliers, ITensorInfo *output_shifts) { ARM_COMPUTE_UNUSED(weights); ARM_COMPUTE_UNUSED(depth_multiplier); 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); unsigned int num_rows_processed_per_iteration = is_stride_1_dilation_1 ? 2 : 1; Window win{}; Status err{}; if(is_data_type_quantized_asymmetric(input->data_type())) { const unsigned int num_elems_accessed_per_iteration = 4; 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 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((output_multipliers != nullptr) && (output_shifts != nullptr)) { AccessWindowHorizontal output_multipliers_access(output_multipliers, 0, num_elems_accessed_per_iteration); AccessWindowHorizontal output_shifts_access(output_shifts, 0, num_elems_accessed_per_iteration); window_changed = window_changed || update_window_and_padding(win, input_access, output_access, output_multipliers_access, output_shifts_access); } else { Status err = ARM_COMPUTE_CREATE_ERROR(ErrorCode::RUNTIME_ERROR, "output_multipliers and output_shifts must be non-nullptr for quantized input"); return std::make_pair(err, win); } if(bias != nullptr) { AccessWindowHorizontal bias_access(bias, 0, num_elems_accessed_per_iteration); window_changed = window_changed || update_window_and_padding(win, bias_access); } err = (window_changed) ? ARM_COMPUTE_CREATE_ERROR(ErrorCode::RUNTIME_ERROR, "Insufficient Padding!") : Status{}; } else { unsigned int num_elems_accessed_per_iteration = adjust_vec_size(4 / input->element_size(), input->dimension(0)); win = calculate_max_window(*output, Steps(num_elems_accessed_per_iteration, num_rows_processed_per_iteration)); } return std::make_pair(err, win); } } // namespace CLDepthwiseConvolutionLayer3x3NHWCKernel::CLDepthwiseConvolutionLayer3x3NHWCKernel() : _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, const ICLTensor *output_multipliers, const ICLTensor *output_shifts) { configure(CLKernelLibrary::get().get_compile_context(), input, weights, biases, output, conv_info, depth_multiplier, act_info, dilation, output_multipliers, output_shifts); } void CLDepthwiseConvolutionLayer3x3NHWCKernel::configure(const CLCompileContext &compile_context, 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, const ICLTensor *output_multipliers, const ICLTensor *output_shifts) { 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, (output_multipliers != nullptr) ? output_multipliers->info() : nullptr, (output_shifts != nullptr) ? output_shifts->info() : nullptr)); auto padding_info = get_padding_info({ input, weights, biases, output }); auto win_config = validate_and_configure_window(input->info(), weights->info(), biases != nullptr ? biases->info() : nullptr, output->info(), conv_info, depth_multiplier, dilation, (output_multipliers != nullptr) ? output_multipliers->info() : nullptr, (output_shifts != nullptr) ? output_shifts->info() : nullptr); 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_quantized_per_channel = is_data_type_quantized_per_channel(weights->info()->data_type()); const bool is_dot8_supported = dot8_supported(CLKernelLibrary::get().get_device()) && !is_quantized_per_channel; _input = input; _output = output; _weights = weights; _biases = biases; _conv_stride_y = conv_info.stride().second; _num_planes_processed_per_iteration = is_stride_1_dilation_1 ? 2 : 1; _output_multipliers = output_multipliers; _output_shifts = output_shifts; _is_quantized = is_data_type_quantized_asymmetric(input->info()->data_type()); if(_is_quantized) { _border_size = BorderSize(input->info()->padding()); // If QASYMM8 and the 8 bit dot product is available, force _num_planes_processed_per_iteration to 1 if(is_dot8_supported) { _num_planes_processed_per_iteration = 1; } } unsigned int num_elems_accessed_per_iteration = _is_quantized ? 4 : adjust_vec_size(4 / input->info()->element_size(), input->info()->dimension(0)); unsigned int num_rows_processed_per_iteration = is_stride_1_dilation_1 ? 2 : 1; CLBuildOptions build_opts; build_opts.add_option("-DDATA_TYPE=" + get_cl_type_from_data_type(_input->info()->data_type())); build_opts.add_option("-DACTIVATION_TYPE=" + lower_string(string_from_activation_func(act_info.activation()))); build_opts.add_option("-DVEC_SIZE=" + support::cpp11::to_string(num_elems_accessed_per_iteration)); build_opts.add_option("-DSRC_DIM_1=" + support::cpp11::to_string(_input->info()->dimension(1))); 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("-DPARTIAL_STORE_N0=" + support::cpp11::to_string(input->info()->dimension(0) % num_elems_accessed_per_iteration)); build_opts.add_option_if(_biases != nullptr, "-DHAS_BIAS"); 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))))); if(_is_quantized) { 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(); 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_if(is_quantized_per_channel, "-DPER_CHANNEL_QUANTIZATION"); build_opts.add_option_if(is_dot8_supported, "-DIS_DOT8"); // Compute non-per-channel multiplier and shift anyway to make OpenCL kernel simpler float multiplier = iq_info.scale * wq_info.scale / oq_info.scale; int output_multiplier = 0; int output_shift = 0; quantization::calculate_quantized_multiplier(multiplier, &output_multiplier, &output_shift); 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()) { int a_val{}; int b_val{}; std::tie(b_val, a_val) = get_quantized_activation_min_max(act_info, input->info()->data_type(), 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)); } build_opts.add_option("-DWEIGHTS_TYPE=" + get_cl_type_from_data_type(weights->info()->data_type())); build_opts.add_option("-DWEIGHTS_PROMOTED_TYPE=" + get_cl_promoted_type_from_data_type(weights->info()->data_type())); } 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())); } 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_1=" + support::cpp11::to_string(_output->info()->dimension(1))); build_opts.add_option("-DDST_DIM_2=" + support::cpp11::to_string(_output->info()->dimension(2))); build_opts.add_option("-DPARTIAL_STORE_M0=" + support::cpp11::to_string((input->info()->dimension(1) + conv_info.pad_left() + conv_info.pad_right()) % num_rows_processed_per_iteration)); } 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("-DDILATION_X=" + support::cpp11::to_string(dilation.x())); build_opts.add_option("-DDILATION_Y=" + support::cpp11::to_string(dilation.y())); } std::string kernel_name; // Create kernel if(_is_quantized) { kernel_name = std::string("dwc_3x3_reshaped_quantized8"); 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" : ""); } ICLKernel::configure_internal(win_config.second); _kernel = create_kernel(compile_context, kernel_name, build_opts.options()); ARM_COMPUTE_ERROR_ON(!_is_quantized && has_padding_changed(padding_info)); // 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, const ITensorInfo *output_multipliers, const ITensorInfo *output_shifts) { ARM_COMPUTE_RETURN_ON_ERROR(validate_arguments(input, weights, biases, output, conv_info, depth_multiplier, act_info, dilation, output_multipliers, output_shifts)); 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, (output_multipliers != nullptr) ? output_multipliers->clone().get() : nullptr, (output_shifts != nullptr) ? output_shifts->clone().get() : nullptr) .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); const size_t total_batches = _input->info()->tensor_shape().total_size_upper(3); Window win = window.collapse_if_possible(ICLKernel::window(), Window::DimZ); win.set(Window::DimZ, Window::Dimension(0, std::ceil(_output->info()->dimension(2) / static_cast(_num_planes_processed_per_iteration)) * total_batches, 1)); unsigned int idx = 2 * num_arguments_per_4D_tensor() + (_is_quantized ? num_arguments_per_2D_tensor() : num_arguments_per_3D_tensor()); if(_is_quantized) { Window slice; slice.use_tensor_dimensions(_output_multipliers->info()->tensor_shape()); slice.set_dimension_step(Window::DimX, window.x().step()); add_1D_tensor_argument(idx, _output_multipliers, slice); add_1D_tensor_argument(idx, _output_shifts, slice); } 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); } if(_is_quantized) { // 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()->dimension(1) * _input->info()->dimension(2)) + (_input->info()->padding().bottom + _input->info()->padding().top) * (_input->info()->dimension( 2) - 1)) * _input->info()->strides_in_bytes().y(); _kernel.setArg(idx, max_offset); } Window slice = win.first_slice_window_4D(); do { unsigned int idx = 0; add_4D_tensor_argument(idx, _input, slice); add_4D_tensor_argument(idx, _output, slice); if(_is_quantized) { add_2D_tensor_argument(idx, _weights, slice); } else { add_3D_tensor_argument(idx, _weights, slice); } enqueue(queue, *this, slice, lws_hint()); } while(win.slide_window_slice_4D(slice)); } } // namespace arm_compute