/* * Copyright (c) 2018-2024 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/cpu/operators/internal/CpuGemmAssemblyDispatch.h" #include "arm_compute/runtime/NEON/NEScheduler.h" #include "src/core/CPP/Validate.h" #include "src/core/helpers/MemoryHelpers.h" #include "src/core/NEON/kernels/arm_gemm/utils.hpp" #include "src/core/utils/AssemblyUtils.h" #include "src/cpu/kernels/assembly/arm_gemm.hpp" #include "src/cpu/kernels/assembly/CpuGemmAssemblyWrapperKernel.h" #include "src/cpu/operators/CpuTranspose.h" #include "src/cpu/utils/CpuAuxTensorHandler.h" #include namespace arm_compute { namespace cpu { namespace { /** Run pretranspose_B_array in parallel (1D static scheduling) * * @tparam TypeInput * @tparam TypeOutput * * @param[in] gemm_asm GemmCommon kernel to run * @param[in] dst Pretransposed B array * @param[in] src B array to be pretransposed * @param[in] src_ld Stride in y * @param[in] src_multi_stride Stride in z ("multi") * @param[in] num_threads Number of threads to run this method. Must be >= 1 */ template void run_parallel_pretranspose_B_array(arm_gemm::GemmCommon *gemm_asm, ITensor *dst, const TypeInput *src, int src_ld, int src_multi_stride, unsigned int num_threads, bool transpose) { ARM_COMPUTE_ERROR_ON(gemm_asm == nullptr); ARM_COMPUTE_ERROR_ON(num_threads == 0); // The window size is also the total workload size const unsigned int wsize = gemm_asm->get_B_pretranspose_window_size(); std::vector workloads(num_threads); for (unsigned int t = 0; t < num_threads; ++t) { workloads[t] = [=](const ThreadInfo &info) { const unsigned int start = (info.thread_id * wsize) / num_threads; const unsigned int end = ((info.thread_id + 1) * wsize) / num_threads; if (start < end) { gemm_asm->pretranspose_B_array_part(dst->buffer(), src, src_ld, src_multi_stride, transpose, start, end); } }; } NEScheduler::get().run_tagged_workloads(workloads, "CpuGemmAssemblyDispatch/pretranspose_B_array"); } } // namespace using namespace arm_compute::experimental; namespace { struct free_delete { void operator()(void *x) { free(x); } }; struct Params { unsigned int M; unsigned int N; unsigned int K; unsigned int batches; unsigned int multis; unsigned int sections; bool indirect; }; Params extract_parameters(const ITensorInfo *a, const ITensorInfo *b, const ITensorInfo *d, const AsmGemmInfo &info) { ARM_COMPUTE_ERROR_ON_NULLPTR(a, b, d); Params p; p.M = d->tensor_shape().y(); p.K = a->tensor_shape().x(); p.N = d->tensor_shape().x(); p.batches = 1; p.multis = 1; p.sections = 1; p.indirect = false; if (info.method == AsmConvMethod::Conv || info.method == AsmConvMethod::Indirect) { p.indirect = true; p.sections = b->tensor_shape()[2] * b->tensor_shape()[3]; } else { p.multis = b->tensor_shape().z(); p.batches = d->tensor_shape().total_size_upper(2) / p.multis; } // Update M in case of GEMM3D for output if (info.depth_output_gemm3d != 0) { p.M = d->tensor_shape().y() * d->tensor_shape().z(); p.batches = d->tensor_shape().total_size_upper(3) / p.multis; } return p; } IScheduler::Hints scheduling_hint_heuristic(arm_gemm::GemmMethod method, DataType data_type) { // Schedule assembly kernel const int granule_threshold = 200; IScheduler::Hints scheduling_hint = IScheduler::Hints(Window::DimX); if (method == arm_gemm::GemmMethod::GEMM_INTERLEAVED && data_type == DataType::F32) { scheduling_hint = IScheduler::Hints(Window::DimX, IScheduler::StrategyHint::DYNAMIC, granule_threshold); } else if (method == arm_gemm::GemmMethod::GEMM_INTERLEAVED_2D && (data_type == DataType::F32 || data_type == DataType::F16 || data_type == DataType::U8 || data_type == DataType::S8)) { //GEMM_INTERLEAVED supports 2D parallelism, IScheduler::split_dimensions_all signals to parallelise over all window dimensions scheduling_hint = IScheduler::Hints(IScheduler::split_dimensions_all, IScheduler::StrategyHint::STATIC, granule_threshold); } else if (method == arm_gemm::GemmMethod::QUANTIZE_WRAPPER_2D && (data_type == DataType::QASYMM8 || data_type == DataType::QASYMM8_SIGNED)) { //special case for QASYMM8 to support 2D parallelism, scheduler here may be tweaked differently compared to FP32 case scheduling_hint = IScheduler::Hints(IScheduler::split_dimensions_all, IScheduler::StrategyHint::STATIC, granule_threshold); } return scheduling_hint; } /** Fallback in case ACL doesn't have a function */ template class Fallback : public CpuGemmAssemblyDispatch::IFallback { public: /** Destructor */ ~Fallback() = default; /** Initialise the functions's input and output. * * @param[in] a Input tensor containing the Matrix A. * @param[in] b Input tensor containing the Matrix B. * @param[in] c Input tensor containing the Matrix C. * @param[out] d Output tensor to store the result of matrix multiplication. * @param[in] args Matrix multiplication information. * @param[in] gemm_info GEMM meta-data * @param[in] os Output stage meta-data. */ void configure(const ITensorInfo *a, const ITensorInfo *b, const ITensorInfo *c, ITensorInfo *d, arm_gemm::GemmArgs args, const AsmGemmInfo &gemm_info, const OutputStage &os = {}); /** Set requantization shifts to be used * * @param[in] shifts Requantization shifts * * @return Pointer to the shift data */ /** Set requantization data to be used * * * @param shifts Requantization shifts * @param multipliers Requantization multipliers * * @return A tuple with the pointers to the shift and multiplier data respectively */ std::tuple set_requantize_data(const std::vector &shifts, const std::vector &multipliers); // Inherited methods overridden: void run(ITensorPack &tensors) override; void prepare(ITensorPack &tensors) override; bool is_configured() const override; experimental::MemoryRequirements workspace() const override; bool isVarWeightsKernel() const override { if (!_gemm_kernel_asm) return false; const arm_compute::WeightFormat wf = assembly_utils::map_to_arm_compute_weight_format(_gemm_kernel_asm->get_config().weight_format); return wf != arm_compute::WeightFormat::UNSPECIFIED && wf != arm_compute::WeightFormat::ANY; } private: enum AuxTensorIdx { AsmGemmWorkspace = 0, PrePretransposedB, /* Transposed B (rhs) before being passed to gemm or pretranspose_B_array */ Pretranspose, Count }; /** Configure the indirect buffer * * @param[in] a Input tensor containing the Matrix A. * @param[in] b Input tensor containing the Matrix B. * @param[out] d Output tensor to store the result of matrix multiplication. * @param[in] info GEMM meta-data */ void configure_indirect(const ITensorInfo *a, const ITensorInfo *b, const ITensorInfo *d, const AsmGemmInfo &info); /** Prepare the indirect buffer */ void prepare_indirect_buffer(ITensorPack &tensors); /** Operator to transpose B before gemm or pretranspose_B_array*/ std::unique_ptr _pre_pretranspose_b{nullptr}; /** Assembly Gemm kernel */ std::shared_ptr> _gemm_kernel_asm{nullptr}; /** Optimised Arm® Neon™ kernel */ std::unique_ptr _optimised_kernel{nullptr}; /** Assembly GEMM workspace tensor info */ TensorInfo _workspace_info{}; /** Pre-pre-transposed B tensor info */ TensorInfo _pre_pretransposed_b_info{}; /** Pre-transpose tensor info */ TensorInfo _pretranspose_info{}; /** Prepared flag */ bool _is_prepared{false}; /** GEMM meta-data */ AsmGemmInfo _gemm_info{}; /** GEMM kernel description */ arm_gemm::KernelDescription _kernel_info{}; /** Per channel quantization shifts */ std::vector _shifts{}; std::vector right_shifts{}; std::vector left_shifts{}; /** Per channel quantization multipliers */ std::vector _multipliers{}; /** Indirect buffer */ std::unique_ptr _indirect_arg{}; std::unique_ptr _indirect_buf{}; std::vector _indirect_pad{}; arm_gemm::ConvolutionParameters _cp{}; experimental::MemoryRequirements _aux_mem{Count}; bool _B_pretranspose_required{false}; bool _is_b_constant{true}; bool _is_c_constant{true}; bool _run_pre_pretranspose_b{false}; bool _B_pre_pretranspose_required{false}; }; template std::tuple Fallback::set_requantize_data(const std::vector &shifts, const std::vector &multipliers) { _multipliers = multipliers; _shifts = shifts; bool need_left = false; for (const auto s : _shifts) { left_shifts.push_back(std::max(-s, int32_t(0))); right_shifts.push_back(std::min(-s, int32_t(0))); if (s < 0 && !need_left) { need_left = true; } } return std::make_tuple(need_left, left_shifts.data(), right_shifts.data(), _multipliers.data()); } template void Fallback::prepare_indirect_buffer(ITensorPack &tensors) { auto a = tensors.get_const_tensor(TensorType::ACL_SRC_0); const TypeInput *A_ptr = reinterpret_cast(a->buffer()); const int multis = 1; const int batches = a->info()->tensor_shape().total_size_upper(3); const size_t stride_A = a->info()->strides_in_bytes().y() / sizeof(TypeInput); const size_t batch_stride_A = a->info()->strides_in_bytes()[3] / sizeof(TypeInput); const size_t multi_stride_A = a->info()->strides_in_bytes()[4] / sizeof(TypeInput); const size_t output_hw = _cp.output_height * _cp.output_width; const int batch_size = _cp.kernel_height * _cp.kernel_width * output_hw * sizeof(TypeInput); const size_t batch_stride = batch_size / sizeof(TypeInput); const int multi_size = batch_size * batches; const size_t multi_stride = multi_size / sizeof(TypeInput); for (int64_t m = 0; m < multis; m++) { for (int64_t b = 0; b < batches; b++) { for (int64_t output_y = 0; output_y < _cp.output_height; output_y++) { for (int64_t output_x = 0; output_x < _cp.output_width; output_x++) { int64_t output_xy = (output_y * _cp.output_width) + output_x; for (int64_t kernel_y = 0; kernel_y < _cp.kernel_height; kernel_y++) { for (int64_t kernel_x = 0; kernel_x < _cp.kernel_width; kernel_x++) { int64_t input_x = (output_x * _cp.output_stride_w) + kernel_x - _cp.padding_left; int64_t input_y = (output_y * _cp.output_stride_h) + kernel_y - _cp.padding_top; int64_t kernel_xy = (kernel_y * _cp.kernel_width) + kernel_x; int64_t input_xy = (input_y * _cp.input_width) + input_x; if (input_x < 0 || input_x >= _cp.input_width || input_y < 0 || input_y >= _cp.input_height) { _indirect_buf .get()[m * multi_stride + b * batch_stride + kernel_xy * output_hw + output_xy] = _indirect_pad.data(); } else { _indirect_buf .get()[m * multi_stride + b * batch_stride + kernel_xy * output_hw + output_xy] = A_ptr + (m * multi_stride_A + b * batch_stride_A + input_xy * stride_A); } } } } } } } } template void Fallback::configure_indirect(const ITensorInfo *a, const ITensorInfo *b, const ITensorInfo *d, const AsmGemmInfo &info) { ARM_COMPUTE_ERROR_ON(!(info.method == AsmConvMethod::Conv || info.method == AsmConvMethod::Indirect)); float zeropad = 0.f; if (is_data_type_quantized(a->data_type())) { zeropad = a->quantization_info().uniform().offset; } const int64_t input_width = static_cast(a->tensor_shape()[1]); const int64_t input_height = static_cast(a->tensor_shape()[2]); const int64_t input_channels = static_cast(a->tensor_shape()[0]); const int64_t kernel_width = static_cast(b->tensor_shape()[2]); const int64_t kernel_height = static_cast(b->tensor_shape()[3]); const int64_t output_width = static_cast(d->tensor_shape()[1]); const int64_t output_height = static_cast(d->tensor_shape()[2]); _cp = {input_width, input_height, input_channels, kernel_width, kernel_height, output_width, output_height, info.ps_info.stride().first, info.ps_info.stride().second, info.padding_top, info.padding_left, zeropad}; if (info.method == AsmConvMethod::Conv) { _gemm_kernel_asm->set_convolution_parameters(_cp); } if (info.method == AsmConvMethod::Indirect) { const unsigned int multis = 1; const unsigned int batches = a->tensor_shape().total_size_upper(3); const unsigned int kernel_hw = _cp.kernel_width * _cp.kernel_height; const unsigned int output_hw = _cp.output_width * _cp.output_height; using TypeInputPtr = TypeInput *; const int batch_size = kernel_hw * output_hw * sizeof(TypeInputPtr); const size_t batch_stride = batch_size / sizeof(TypeInputPtr); const int multi_size = batch_size * batches; const size_t multi_stride = multi_size / sizeof(TypeInputPtr); _indirect_buf = std::unique_ptr( reinterpret_cast(malloc(multi_size * multis))); _indirect_arg = std::unique_ptr( reinterpret_cast(malloc(sizeof(TypeInput **) * kernel_hw * multis * batches))); _indirect_pad = std::vector(_cp.input_channels, TypeInput(zeropad)); // Set indirect argument int64_t pos = 0; for (int64_t m = 0; m < multis; m++) { for (int64_t b = 0; b < batches; b++) { for (int64_t kernel_xy = 0; kernel_xy < kernel_hw; kernel_xy++) { (_indirect_arg.get())[pos++] = _indirect_buf.get() + m * multi_stride + b * batch_stride + kernel_xy * output_hw; } } } _gemm_kernel_asm->set_indirect_parameters(a->tensor_shape()[0], _indirect_arg.get()); } } template void Fallback::configure(const ITensorInfo *a, const ITensorInfo *b, const ITensorInfo *c, ITensorInfo *d, arm_gemm::GemmArgs args, const AsmGemmInfo &gemm_info, const OutputStage &os) { _is_b_constant = b->are_values_constant(); _is_c_constant = c ? c->are_values_constant() : true; _gemm_kernel_asm = arm_gemm::gemm(args, os); if (_gemm_kernel_asm == nullptr) { //configuration not supported: Leave function unconfigured: return; } arm_gemm::GemmConfig gemm_cfg = _gemm_kernel_asm->get_config(); // arm_compute wrapper for the Gemm object (see above) auto acl_gemm_wrapper = std::make_unique>(); ARM_COMPUTE_ERROR_ON(acl_gemm_wrapper == nullptr); acl_gemm_wrapper->configure(_gemm_kernel_asm.get(), gemm_cfg.filter); const size_t workspace_size = _gemm_kernel_asm->get_working_size(); const unsigned int alignment = 4096; _workspace_info = TensorInfo(TensorShape(workspace_size), 1, DataType::U8); _aux_mem[AsmGemmWorkspace] = MemoryInfo(offset_int_vec(AsmGemmWorkspace), MemoryLifetime::Temporary, workspace_size, alignment); //if we disable this code below in brackets then ConvLayer deadlocks when threads > 1 and //the shapes are In=1x1x1024 Weights=1x1x1024x1001 Biases=1001 Out=1x1x1001 { const unsigned int window_size = _gemm_kernel_asm->get_window_size().total_size(); if (window_size < static_cast(args._maxthreads)) { _gemm_kernel_asm->set_nthreads(window_size); } } _optimised_kernel = std::move(acl_gemm_wrapper); _gemm_info = gemm_info; // Check if we need to pre-pretranspose B. Fixed format kernels need no pre-pretranspose. _B_pre_pretranspose_required = _gemm_info.transpose_b && !isVarWeightsKernel(); _B_pretranspose_required = _gemm_kernel_asm->B_pretranspose_required(); const bool kernel_supports_transpose = _gemm_kernel_asm->B_pretranspose_supports_transpose(); const bool kernel_can_fuse_transpose = _B_pretranspose_required && kernel_supports_transpose; _run_pre_pretranspose_b = _B_pre_pretranspose_required && !kernel_can_fuse_transpose; if (_run_pre_pretranspose_b) { _pre_pretranspose_b = std::make_unique(); _pre_pretranspose_b->configure(b, &_pre_pretransposed_b_info); MemoryLifetime lifetime; if (_is_b_constant) { if (_B_pretranspose_required) { // PrePretransposedB tensor is only used in prepare(), but is then succeeded by Pretranspose // So PrePretransposedB can be freed inside prepare() lifetime = MemoryLifetime::Prepare; } else { // PrePretransposedB tensor is only used in prepare(), but is the final transformation of B // So PrePretransposedB needs to persist beyond prepare() lifetime = MemoryLifetime::Persistent; } } else { // PrePretransposedB tensor is always used in run() and doesn't need to persist lifetime = MemoryLifetime::Temporary; } // Forcing 128-byte alignment (required by 32-bit kernels) const unsigned int alignment = 128; _aux_mem[PrePretransposedB] = MemoryInfo(offset_int_vec(PrePretransposedB), lifetime, _pre_pretransposed_b_info.total_size(), alignment); } // Check for pre-transposed support if (_B_pretranspose_required) { // Fixed format kernels need no pretranspose. ARM_COMPUTE_ERROR_ON(arm_compute::is_fixed_format( assembly_utils::map_to_arm_compute_weight_format(_gemm_kernel_asm->get_config().weight_format))); // Forcing 128-byte alignment (required by 32-bit kernels) const unsigned int alignment = 128; const size_t B_pretranspose_size = _gemm_kernel_asm->get_B_pretransposed_array_size(); _pretranspose_info = TensorInfo(TensorShape(B_pretranspose_size), 1, DataType::U8); _aux_mem[Pretranspose] = MemoryInfo(offset_int_vec(Pretranspose), MemoryLifetime::Persistent, B_pretranspose_size, alignment); } // Handle indirect GEMM convolution if (gemm_info.method == AsmConvMethod::Conv || gemm_info.method == AsmConvMethod::Indirect) { configure_indirect(a, b, d, gemm_info); } if (std::is_same::value) { // Output dequantization is just the two src scales multiplied together _gemm_kernel_asm->set_dequantize_scale(a->quantization_info().uniform().scale * b->quantization_info().uniform().scale); } } template void Fallback::prepare(ITensorPack &tensors) { if (!_is_prepared) { auto b = tensors.get_const_tensor(TensorType::ACL_SRC_1); auto c = tensors.get_const_tensor(TensorType::ACL_SRC_2); ARM_COMPUTE_ERROR_ON_NULLPTR(b); // Setup up matrix bias in the assembly kernel, it's just a pointer to matrix C. if (c && c->info()->data_type() == DataType::S32) { _gemm_kernel_asm->set_quantized_bias( reinterpret_cast(c->buffer() + c->info()->offset_first_element_in_bytes()), 0); } const ITensor *b_to_use = b; // Pre-pretranspose B if required CpuAuxTensorHandler pre_pretransposed_b( offset_int_vec(PrePretransposedB), _pre_pretransposed_b_info, tensors, /*pack_inject: no need to inject into tensors*/ false, /*bypass_alloc: no need to allocate if pre-pretranspose B is not required as this handle will not be used*/ !_run_pre_pretranspose_b); if (_run_pre_pretranspose_b) { ARM_COMPUTE_ERROR_ON(_pre_pretranspose_b == nullptr); ITensorPack pre_pretranspose_pack{{ACL_SRC, b_to_use}, {ACL_DST, pre_pretransposed_b.get()}}; _pre_pretranspose_b->run(pre_pretranspose_pack); b_to_use = pre_pretransposed_b.get(); } // Pretranspose B if required if (_B_pretranspose_required) { // Fixed format kernels need no pretranspose. ARM_COMPUTE_ERROR_ON(arm_compute::is_fixed_format( assembly_utils::map_to_arm_compute_weight_format(_gemm_kernel_asm->get_config().weight_format))); const int ldb = b_to_use->info()->strides_in_bytes().y() / b_to_use->info()->element_size(); const auto in1_ptr = reinterpret_cast(b_to_use->buffer() + b_to_use->info()->offset_first_element_in_bytes()); const int multi_stride_b = b_to_use->info()->strides_in_bytes().z() / b_to_use->info()->element_size(); CpuAuxTensorHandler pretranspose(offset_int_vec(Pretranspose), _pretranspose_info, tensors, false); ARM_COMPUTE_ERROR_ON(pretranspose.get()->buffer() == nullptr); const bool kernel_supports_transpose = _gemm_kernel_asm->B_pretranspose_supports_transpose(); run_parallel_pretranspose_B_array( _gemm_kernel_asm.get(), pretranspose.get(), in1_ptr, ldb, multi_stride_b, NEScheduler::get().num_threads(), _B_pre_pretranspose_required && kernel_supports_transpose); b->mark_as_unused(); // Note that we don't need to mark b_to_use as unused, as if it's been assigned to pre_pretransposed_b, // its memory will be auto-managed by the handler } if (_gemm_info.method == AsmConvMethod::Indirect) { prepare_indirect_buffer(tensors); } _is_prepared = true; } } template bool Fallback::is_configured() const { return _optimised_kernel != nullptr; } template experimental::MemoryRequirements Fallback::workspace() const { return _aux_mem; } template void Fallback::run(ITensorPack &tensors) { auto a = tensors.get_const_tensor(TensorType::ACL_SRC_0); auto b = tensors.get_const_tensor(TensorType::ACL_SRC_1); auto c = tensors.get_const_tensor(TensorType::ACL_SRC_2); auto d = tensors.get_tensor(TensorType::ACL_DST); ARM_COMPUTE_ERROR_ON_NULLPTR(a, d); // Only update at runtime if the src quantization is dynamic if (std::is_same::value && (a->info()->quantization_info().is_dynamic() || b->info()->quantization_info().is_dynamic())) { // Output dequantization is just the two src scales multiplied together _gemm_kernel_asm->set_dequantize_scale(a->info()->quantization_info().uniform().scale * b->info()->quantization_info().uniform().scale); } int lda = a->info()->strides_in_bytes().y() / a->info()->element_size(); int ldb = 0; const int ldd = d->info()->strides_in_bytes().y() / d->info()->element_size(); const size_t a_batch_idx = _gemm_info.reinterpret_input_as_3d != 0 ? 3 : 2; const size_t a_multi_idx = a_batch_idx + 1; const size_t d_batch_idx = _gemm_info.depth_output_gemm3d != 0 ? 3 : 2; const size_t d_multi_idx = d_batch_idx + 1; int batch_stride_a = a->info()->strides_in_bytes()[a_batch_idx] / a->info()->element_size(); const int batch_stride_d = d->info()->strides_in_bytes()[d_batch_idx] / d->info()->element_size(); int multi_stride_a = a->info()->strides_in_bytes()[a_multi_idx] / a->info()->element_size(); int multi_stride_b = 0; const int multi_stride_d = d->info()->strides_in_bytes()[d_multi_idx] / d->info()->element_size(); auto in0_ptr = reinterpret_cast(a->buffer() + a->info()->offset_first_element_in_bytes()); const TypeInput *in1_ptr = nullptr; auto out_ptr = reinterpret_cast(d->buffer() + d->info()->offset_first_element_in_bytes()); const ITensor *b_to_use = b; // Pre-pretranspose B if required CpuAuxTensorHandler pre_pretransposed_b( offset_int_vec(PrePretransposedB), _pre_pretransposed_b_info, tensors, false /*pack_inject: no need to inject into tensors*/, !_run_pre_pretranspose_b /*bypass_alloc: no need to allocate if pre-pretranspose B is not required as this handle will not be used*/); if (b_to_use && !_is_b_constant && _run_pre_pretranspose_b) { ARM_COMPUTE_ERROR_ON(_pre_pretranspose_b == nullptr); ITensorPack pre_pretranspose_pack{{ACL_SRC, b_to_use}, {ACL_DST, pre_pretransposed_b.get()}}; _pre_pretranspose_b->run(pre_pretranspose_pack); b_to_use = pre_pretransposed_b.get(); } // Check if B is pre-tranposed and de-reference if not if (b_to_use && !_gemm_kernel_asm->B_is_pretransposed()) { ldb = b_to_use->info()->strides_in_bytes().y() / b_to_use->info()->element_size(); multi_stride_b = b_to_use->info()->strides_in_bytes().z() / b_to_use->info()->element_size(); in1_ptr = reinterpret_cast(b_to_use->buffer() + b_to_use->info()->offset_first_element_in_bytes()); } // If necessary, run pretranspose every time if either weights or biases are non-constant if ((b_to_use && !_is_b_constant) || (c && !_is_c_constant && c->info()->data_type() == DataType::S32)) { if (c && c->info()->data_type() == DataType::S32) { _gemm_kernel_asm->set_quantized_bias( reinterpret_cast(c->buffer() + c->info()->offset_first_element_in_bytes()), 0); } // Pretranspose B if required if (b_to_use && _B_pretranspose_required) { // Fixed format kernels need no pretranspose. ARM_COMPUTE_ERROR_ON(arm_compute::is_fixed_format( assembly_utils::map_to_arm_compute_weight_format(_gemm_kernel_asm->get_config().weight_format))); const int ldb = b_to_use->info()->strides_in_bytes().y() / b_to_use->info()->element_size(); const auto b_ptr = reinterpret_cast(b_to_use->buffer() + b_to_use->info()->offset_first_element_in_bytes()); const int multi_stride_b = b_to_use->info()->strides_in_bytes().z() / b_to_use->info()->element_size(); CpuAuxTensorHandler pretranspose(offset_int_vec(Pretranspose), _pretranspose_info, tensors, true); ARM_COMPUTE_ERROR_ON(pretranspose.get()->buffer() == nullptr); if (_is_b_constant) { _gemm_kernel_asm->requantize_bias(pretranspose.get()->buffer(), b_ptr, ldb, multi_stride_b); } else { const bool kernel_supports_transpose = _gemm_kernel_asm->B_pretranspose_supports_transpose(); run_parallel_pretranspose_B_array( _gemm_kernel_asm.get(), pretranspose.get(), b_ptr, ldb, multi_stride_b, NEScheduler::get().num_threads(), _B_pre_pretranspose_required && kernel_supports_transpose); } } } const auto scheduling_hint = scheduling_hint_heuristic(_kernel_info.method, d->info()->data_type()); // Set workspace if needed and reset number of threads as buffer manager gets re-created with max_threads CpuAuxTensorHandler workspace(offset_int_vec(AsmGemmWorkspace), _workspace_info, tensors, false); if (workspace.get()->buffer() != nullptr) { _gemm_kernel_asm->set_working_space(reinterpret_cast(workspace.get()->buffer())); const unsigned int split_dim = scheduling_hint.split_dimension(); const unsigned int window_size = _gemm_kernel_asm->get_window_size().total_size(); unsigned int num_threads = NEScheduler::get().num_threads(); if (window_size < num_threads) { num_threads = window_size; } if (split_dim != IScheduler::split_dimensions_all) { // Make sure the kernel does not expect more threads than we can actually spawn const unsigned int num_iterations = _optimised_kernel.get()->window().num_iterations(split_dim); num_threads = std::min(num_iterations, num_threads); } _gemm_kernel_asm->set_nthreads(num_threads); } // Prepare assembly kernel prepare(tensors); // Setup up matrix bias in the assembly kernel, it's just a pointer to matrix C. TypeOutput *bias = nullptr; if (c && c->info()->data_type() != DataType::S32) { bias = reinterpret_cast(c->buffer() + c->info()->offset_first_element_in_bytes()); } if (_gemm_info.method == AsmConvMethod::Indirect) { in0_ptr = nullptr; lda = 0; batch_stride_a = 0; multi_stride_a = 0; } // Set gemm parameters _gemm_kernel_asm->set_arrays(in0_ptr, lda, batch_stride_a, multi_stride_a, in1_ptr, ldb, multi_stride_b, out_ptr, ldd, batch_stride_d, multi_stride_d, bias, 0); // Schedule NEScheduler::get().schedule(_optimised_kernel.get(), scheduling_hint); } template void create_arm_gemm(std::unique_ptr &arm_gemm, const ITensorInfo *a, const ITensorInfo *b, const ITensorInfo *c, ITensorInfo *d, arm_gemm::Activation activation, const AsmGemmInfo &info) { Params p = extract_parameters(a, b, d, info); const CPUInfo &ci = NEScheduler::get().cpu_info(); unsigned int num_threads = NEScheduler::get().num_threads(); arm_gemm::GemmConfig cfg; cfg.weight_format = assembly_utils::map_to_arm_gemm_weight_format(info.weight_format); arm_gemm::GemmArgs args(&ci, p.M, p.N, p.K, p.sections, p.batches, p.multis, p.indirect, activation, num_threads, info.fixed_format, info.fast_mode, info.accumulate, &cfg); // Create arm_gemm fallback auto fallback = std::make_unique>(); fallback->configure(a, b, c, d, args, info); arm_gemm = std::move(fallback); } template void create_arm_gemm_dequant(std::unique_ptr &arm_gemm, const ITensorInfo *a, const ITensorInfo *b, const ITensorInfo *c, ITensorInfo *d, arm_gemm::Activation activation, const AsmGemmInfo &info) { ARM_COMPUTE_UNUSED(activation); Params p = extract_parameters(a, b, d, info); const CPUInfo &ci = NEScheduler::get().cpu_info(); const unsigned int num_threads = NEScheduler::get().num_threads(); arm_gemm::GemmConfig cfg; cfg.weight_format = assembly_utils::map_to_arm_gemm_weight_format(info.weight_format); arm_gemm::GemmArgs args(&ci, p.M, p.N, p.K, p.sections, p.batches, p.multis, p.indirect, activation, num_threads, info.fixed_format, info.fast_mode, info.accumulate, &cfg); // Create arm_gemm fallback auto fallback = std::make_unique>(); // Configure requantization info const GEMMLowpOutputStageInfo os_info = info.output_stage; arm_gemm::DequantizeFloat gemm_dequant_info{}; gemm_dequant_info = arm_gemm::DequantizeFloat(d->quantization_info().uniform().scale); fallback->configure(a, b, c, d, args, info, gemm_dequant_info); arm_gemm = std::move(fallback); } template void create_arm_gemm_quant(std::unique_ptr &arm_gemm, const ITensorInfo *a, const ITensorInfo *b, const ITensorInfo *c, ITensorInfo *d, arm_gemm::Activation activation, const AsmGemmInfo &info) { ARM_COMPUTE_UNUSED(activation); Params p = extract_parameters(a, b, d, info); const CPUInfo &ci = NEScheduler::get().cpu_info(); const unsigned int num_threads = NEScheduler::get().num_threads(); arm_gemm::GemmConfig cfg; cfg.weight_format = assembly_utils::map_to_arm_gemm_weight_format(info.weight_format); arm_gemm::GemmArgs args(&ci, p.M, p.N, p.K, p.sections, p.batches, p.multis, p.indirect, activation, num_threads, info.fixed_format, info.fast_mode, info.accumulate, &cfg); // Create arm_gemm fallback auto fallback = std::make_unique>(); // Configure requantization info const int32_t negation = info.negated_offsets ? 1 : -1; const int32_t a_offset = -a->quantization_info().uniform().offset * negation; const int32_t b_offset = -b->quantization_info().uniform().offset * negation; const GEMMLowpOutputStageInfo os_info = info.output_stage; arm_gemm::Requantize32 gemm_requant_info{}; if (os_info.gemmlowp_shifts.size() > 1) { const auto requantize_data = fallback->set_requantize_data(os_info.gemmlowp_shifts, os_info.gemmlowp_multipliers); gemm_requant_info = arm_gemm::Requantize32( nullptr, 0, a_offset, b_offset, os_info.gemmlowp_offset, (std::get<0>(requantize_data)) ? std::get<1>(requantize_data) : nullptr, std::get<2>(requantize_data), std::get<3>(requantize_data), os_info.gemmlowp_min_bound, os_info.gemmlowp_max_bound); } else { gemm_requant_info = arm_gemm::Requantize32(nullptr, 0, a_offset, b_offset, os_info.gemmlowp_offset, -os_info.gemmlowp_shift, os_info.gemmlowp_multiplier, os_info.gemmlowp_min_bound, os_info.gemmlowp_max_bound); } // Configure fallback fallback->configure(a, b, c, d, args, info, gemm_requant_info); arm_gemm = std::move(fallback); } } //namespace CpuGemmAssemblyDispatch::CpuGemmAssemblyDispatch() : _arm_gemm(nullptr) { } Status CpuGemmAssemblyDispatch::has_opt_impl(arm_compute::WeightFormat &expected_weight_format, const ITensorInfo *a, const ITensorInfo *b, const ITensorInfo *c, const ITensorInfo *d, const AsmGemmInfo &info) { ARM_COMPUTE_ERROR_ON_NULLPTR(a, b, d); ARM_COMPUTE_UNUSED(c); arm_gemm::Activation act = assembly_utils::map_to_arm_gemm_activation(info.activation_info); Params p = extract_parameters(a, b, d, info); const CPUInfo &ci = NEScheduler::get().cpu_info(); unsigned int num_threads = NEScheduler::get().num_threads(); arm_gemm::GemmConfig cfg; cfg.weight_format = assembly_utils::map_to_arm_gemm_weight_format(info.weight_format); arm_gemm::WeightFormat arm_gemm_expected_wf = assembly_utils::map_to_arm_gemm_weight_format(expected_weight_format); arm_gemm::GemmArgs args(&ci, p.M, p.N, p.K, p.sections, p.batches, p.multis, p.indirect, act, num_threads, info.fixed_format, info.fast_mode, info.accumulate, &cfg); // TODO: Incorporate info.transpose_b COMPMID-6595 switch (a->data_type()) { case DataType::F32: ARM_COMPUTE_RETURN_ERROR_ON_MSG( !(arm_gemm::has_opt_gemm(arm_gemm_expected_wf, args, {})), "We could not find an optimized kernel for F32 input"); break; #ifdef __aarch64__ case DataType::U8: case DataType::QASYMM8: if (d->data_type() == DataType::S32) { ARM_COMPUTE_RETURN_ERROR_ON_MSG( !(arm_gemm::has_opt_gemm(arm_gemm_expected_wf, args, {})), "We could not find an optimized kernel for U8/QASYMM8 input and U32 output"); } else { ARM_COMPUTE_RETURN_ERROR_ON_MSG( !(arm_gemm::has_opt_gemm(arm_gemm_expected_wf, args, {})), "We could not find an optimized kernel for U8 input and U8 output"); } break; case DataType::S8: case DataType::QASYMM8_SIGNED: if (d->data_type() == DataType::S32) { ARM_COMPUTE_RETURN_ERROR_ON_MSG( !(arm_gemm::has_opt_gemm(arm_gemm_expected_wf, args, {})), "We could not find an optimized kernel for S8/QASYMM8_SIGNED input and S32 output"); } else { ARM_COMPUTE_RETURN_ERROR_ON_MSG( !(arm_gemm::has_opt_gemm(arm_gemm_expected_wf, args, {})), "We could not find an optimized kernel for S8 input and S8 output"); } break; #endif /* __aarch64__ */ #if defined(ARM_COMPUTE_ENABLE_BF16) case DataType::BFLOAT16: { if (d->data_type() == DataType::BFLOAT16) { ARM_COMPUTE_RETURN_ERROR_ON_MSG( !(arm_gemm::has_opt_gemm(arm_gemm_expected_wf, args, {})), "We could not find an optimized kernel for BFLOAT16 input and BFLOAT16 output"); } else { ARM_COMPUTE_RETURN_ERROR_ON_MSG( !(arm_gemm::has_opt_gemm(arm_gemm_expected_wf, args, {})), "We could not find an optimized kernel for BFLOAT16 input and F32 output"); } break; } #endif /* defined(ARM_COMPUTE_ENABLE_BF16) */ #ifdef __ARM_FEATURE_FP16_VECTOR_ARITHMETIC case DataType::F16: ARM_COMPUTE_RETURN_ERROR_ON_MSG( !(arm_gemm::has_opt_gemm(arm_gemm_expected_wf, args, {})), "We could not find an optimized kernel for F16 input and F16 output"); break; #endif /* __ARM_FEATURE_FP16_VECTOR_ARITHMETIC */ default: ARM_COMPUTE_RETURN_ERROR_ON_MSG(true, "Usupported type. Could not find a kernel"); break; } expected_weight_format = assembly_utils::map_to_arm_compute_weight_format(arm_gemm_expected_wf); return Status{}; } Status CpuGemmAssemblyDispatch::validate( const ITensorInfo *a, const ITensorInfo *b, const ITensorInfo *c, const ITensorInfo *d, const AsmGemmInfo &info) { ARM_COMPUTE_UNUSED(c, info); ARM_COMPUTE_RETURN_ERROR_ON_NULLPTR(a, b, d); ARM_COMPUTE_RETURN_ERROR_ON_CPU_F16_UNSUPPORTED(a); ARM_COMPUTE_RETURN_ERROR_ON_CPU_BF16_UNSUPPORTED(a); ARM_COMPUTE_RETURN_ERROR_ON_MSG(!(info.reshape_b_only_on_first_run), "Assembly kernel will not be executed when reshape_b_only_on_first_run is false"); #ifndef __aarch64__ ARM_COMPUTE_RETURN_ERROR_ON_MSG(a->element_size() == 1, "8bit integer types only supported for aarch64"); #endif /* __aarch64__ */ ARM_COMPUTE_RETURN_ERROR_ON_DATA_TYPE_CHANNEL_NOT_IN(a, 1, DataType::U8, DataType::QASYMM8, DataType::QASYMM8_SIGNED, DataType::S8, DataType::BFLOAT16, DataType::F16, DataType::F32); ARM_COMPUTE_RETURN_ERROR_ON_DATA_TYPE_CHANNEL_NOT_IN( b, 1, DataType::U8, DataType::QASYMM8, DataType::QASYMM8_SIGNED, DataType::QSYMM8_PER_CHANNEL, DataType::S8, DataType::BFLOAT16, DataType::F16, DataType::F32); if (is_data_type_quantized_per_channel(b->data_type())) { ARM_COMPUTE_RETURN_ERROR_ON_DATA_TYPE_CHANNEL_NOT_IN(a, 1, DataType::QASYMM8_SIGNED, DataType::S8); } else if (is_fixed_format_fast_math(info.weight_format)) { ARM_COMPUTE_RETURN_ERROR_ON_DATA_TYPE_NOT_IN(a, DataType::F32); ARM_COMPUTE_RETURN_ERROR_ON_DATA_TYPE_NOT_IN(b, DataType::BFLOAT16); } else { ARM_COMPUTE_RETURN_ERROR_ON_MISMATCHING_DATA_TYPES(a, b); } ARM_COMPUTE_RETURN_ERROR_ON_MSG(a->data_type() == DataType::F32 && d->data_type() != DataType::F32, "Only F32 output supported for F32 input"); ARM_COMPUTE_RETURN_ERROR_ON_MSG(a->data_type() == DataType::F16 && d->data_type() != DataType::F16, "Only F16 output supported for F16 input"); ARM_COMPUTE_RETURN_ERROR_ON_MSG(a->data_type() == DataType::BFLOAT16 && (d->data_type() != DataType::F32 && d->data_type() != DataType::BFLOAT16), "Only F32/BFLOAT16 output supported for BFLOAT16 input"); ARM_COMPUTE_RETURN_ERROR_ON_MSG(a->data_type() == DataType::U8 && d->data_type() != DataType::U32, "Only U32 output supported for U8 input"); ARM_COMPUTE_RETURN_ERROR_ON_MSG(a->data_type() == DataType::S8 && d->data_type() != DataType::S32, "Only S32 output supported for S8 input"); ARM_COMPUTE_RETURN_ERROR_ON_MSG(a->data_type() == DataType::QASYMM8 && (d->data_type() != DataType::QASYMM8 && d->data_type() != DataType::S32), "Only QASYMM8/S32 output supported for QASYMM8 input"); arm_compute::WeightFormat expected_weight_format = arm_compute::WeightFormat::UNSPECIFIED; const Status ret = CpuGemmAssemblyDispatch::has_opt_impl(expected_weight_format, a, b, c, d, info); if ((bool)ret && expected_weight_format != arm_compute::WeightFormat::ANY) { // Correctness check: if the format expected by the kernel is // not "any", make sure that the one found matches the format // intended by the caller. ARM_COMPUTE_RETURN_ERROR_ON_MSG( (expected_weight_format != info.weight_format), "The format expected by the kernel does not correspond with the one requested by the user."); } return ret; } bool CpuGemmAssemblyDispatch::is_activation_supported(const ActivationLayerInfo &activation) { arm_gemm::Activation act = assembly_utils::map_to_arm_gemm_activation(activation); return act.type != arm_gemm::Activation::Type::None; } void CpuGemmAssemblyDispatch::configure( const ITensorInfo *a, const ITensorInfo *b, const ITensorInfo *c, ITensorInfo *d, const AsmGemmInfo &info) { ARM_COMPUTE_ERROR_ON_NULLPTR(a, b, d); arm_gemm::Activation act = assembly_utils::map_to_arm_gemm_activation(info.activation_info); //If we don't support a combination of data types, silently return: it is the caller's responsibility to check if configure() was successful via is_configured() if (!CpuGemmAssemblyDispatch::validate(a, b, c, d, info)) { return; } switch (a->data_type()) { case DataType::F32: create_arm_gemm(_arm_gemm, a, b, c, d, act, info); break; #ifdef __aarch64__ case DataType::U8: case DataType::QASYMM8: if (d->data_type() == DataType::S32) { create_arm_gemm(_arm_gemm, a, b, c, d, act, info); } else { create_arm_gemm_quant(_arm_gemm, a, b, c, d, act, info); } break; case DataType::S8: case DataType::QASYMM8_SIGNED: if (d->data_type() == DataType::S32) { create_arm_gemm(_arm_gemm, a, b, c, d, act, info); } else if (d->data_type() == DataType::F32) { create_arm_gemm_dequant(_arm_gemm, a, b, c, d, act, info); } else { create_arm_gemm_quant(_arm_gemm, a, b, c, d, act, info); } break; #endif /* __aarch64__ */ #if defined(ARM_COMPUTE_ENABLE_BF16) case DataType::BFLOAT16: if (d->data_type() == DataType::BFLOAT16) { create_arm_gemm(_arm_gemm, a, b, c, d, act, info); } else { create_arm_gemm(_arm_gemm, a, b, c, d, act, info); } break; #endif /* defined(ARM_COMPUTE_ENABLE_BF16) */ #ifdef __ARM_FEATURE_FP16_VECTOR_ARITHMETIC case DataType::F16: create_arm_gemm(_arm_gemm, a, b, c, d, act, info); break; #endif /* __ARM_FEATURE_FP16_VECTOR_ARITHMETIC */ default: break; } } void CpuGemmAssemblyDispatch::prepare(ITensorPack &tensors) { ARM_COMPUTE_ERROR_ON(_arm_gemm == nullptr); _arm_gemm->prepare(tensors); } bool CpuGemmAssemblyDispatch::is_configured() const { return _arm_gemm && _arm_gemm->is_configured(); } void CpuGemmAssemblyDispatch::run(ITensorPack &tensors) { ARM_COMPUTE_ERROR_ON(_arm_gemm == nullptr); _arm_gemm->run(tensors); } experimental::MemoryRequirements CpuGemmAssemblyDispatch::workspace() const { ARM_COMPUTE_ERROR_ON(_arm_gemm == nullptr); return _arm_gemm->workspace(); } } // namespace cpu } // namespace arm_compute