CPU Direct Execution: Implement Broadcast

src/ngraph/runtime/cpu/builder/broadcast.cpp

+/*******************************************************************************
+* Copyright 2018 Intel Corporation
+*
+* Licensed under the Apache License, Version 2.0 (the "License");
+* you may not use this file except in compliance with the License.
+* You may obtain a copy of the License at
+*
+*     http://www.apache.org/licenses/LICENSE-2.0
+*
+* Unless required by applicable law or agreed to in writing, software
+* distributed under the License is distributed on an "AS IS" BASIS,
+* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+* See the License for the specific language governing permissions and
+* limitations under the License.
+*******************************************************************************/
+
+#include <cstring>
+
+#include "ngraph/op/broadcast.hpp"
+#include "ngraph/runtime/cpu/cpu_builder.hpp"
+#include "ngraph/runtime/cpu/kernel/broadcast.hpp"
+
+using namespace std;
+using namespace ngraph;
+
+namespace ngraph
+{
+    namespace runtime
+    {
+        namespace cpu
+        {
+            template <>
+            void Builder::BUILDER_DECL(ngraph::op::Broadcast)
+            {
+                auto& functors = external_function->get_functors();
+                auto& tensor_data = external_function->get_tensor_data();
+
+                auto& arg_tensor = tensor_data[args[0].get_name()];
+                auto& out_tensor = tensor_data[out[0].get_name()];
+
+                auto arg_shape = args[0].get_shape();
+                auto out_shape = out[0].get_shape();
+                auto arg_rank = arg_shape.size();
+                auto out_rank = out_shape.size();
+
+                auto broadcast = static_cast<const ngraph::op::Broadcast*>(node);
+
+                if (broadcast->get_broadcast_axes().empty())
+                {
+                    size_t size = out[0].get_size() * out[0].get_element_type().size();
+                    auto functor = [&, size](CPURuntimeContext* ctx) {
+                        memcpy(out_tensor, arg_tensor, size);
+                    };
+                    functors.emplace_back(functor);
+                    return;
+                }
+
+                if (!arg_rank)
+                {
+                    arg_rank = 1;
+                    arg_shape = Shape{1};
+                }
+                auto new_shape = Shape(out_rank, 1);
+                const auto& broadcast_axes = broadcast->get_broadcast_axes();
+                size_t i = 0;
+                for (size_t j = 0; j < out_rank; j++)
+                {
+                    if (broadcast_axes.count(j))
+                    {
+                        new_shape[j] = 1;
+                    }
+                    else
+                    {
+                        new_shape[j] = arg_shape[i++];
+                    }
+                }
+
+                std::function<decltype(runtime::cpu::kernel::broadcast<float, 2>)> kernel;
+
+                SELECT_KERNEL_BY_RANK(
+                    kernel, args[0].get_element_type(), out_rank, runtime::cpu::kernel::broadcast);
+
+                auto functor = [&, kernel, new_shape, out_shape](CPURuntimeContext* ctx) {
+                    kernel(arg_tensor, out_tensor, new_shape, out_shape);
+                };
+                functors.emplace_back(functor);
+            }
+
+            REGISTER_OP_BUILDER(Broadcast);
+        }
+    }
+}

src/ngraph/runtime/cpu/kernel/broadcast.hpp

@@ -19,8 +19,10 @@
 #define EIGEN_USE_THREADS
 #include <unsupported/Eigen/CXX11/Tensor>
 
+#include "ngraph/axis_set.hpp"
 #include "ngraph/runtime/cpu/kernel/eigen_thread_pool.hpp"
 #include "ngraph/runtime/reference/broadcast.hpp"
+#include "ngraph/shape.hpp"
 
 namespace ngraph
 {
@@ -30,18 +32,33 @@ namespace ngraph
         {
             namespace kernel
             {
-                template <typename ElementType>
-                void broadcast(void* input0,
+                template <typename ElementType, unsigned int Rank>
+                void broadcast(void* input,
                                void* output,
-                               const Shape& arg0_shape,
-                               const Shape& result_shape,
-                               const AxisSet& broadcast_axes)
+                               const Shape& input_shape,
+                               const Shape& output_shape)
                 {
-                    reference::broadcast<ElementType>(static_cast<const ElementType*>(input0),
-                                                      static_cast<ElementType*>(output),
-                                                      arg0_shape,
-                                                      result_shape,
-                                                      broadcast_axes);
+                    Eigen::array<Eigen::Index, Rank> out_dims;
+                    Eigen::array<Eigen::Index, Rank> in_dims;
+
+                    for (int i = 0; i < Rank; i++)
+                    {
+                        out_dims[i] = output_shape[i];
+                        in_dims[i] = input_shape[i];
+                    }
+
+                    Eigen::TensorMap<Eigen::Tensor<ElementType, Rank, Eigen::RowMajor>> out(
+                        static_cast<ElementType*>(output), out_dims);
+                    Eigen::TensorMap<Eigen::Tensor<ElementType, Rank, Eigen::RowMajor>> in(
+                        static_cast<ElementType*>(input), in_dims);
+
+                    Eigen::array<ptrdiff_t, Rank> factors;
+                    for (int i = 0; i < Rank; i++)
+                    {
+                        factors[i] = output_shape[i] / input_shape[i];
+                    }
+
+                    out.device(eigen::global_thread_pool_device) = in.broadcast(factors);
                 }
             }
         }