Added DEX execution support for ReluBprop (#1305)

src/ngraph/runtime/cpu/CMakeLists.txt

@@ -40,6 +40,7 @@ set(SRC
     builder/max.cpp
     builder/max_pool.cpp
     builder/min.cpp
+    builder/relu.cpp
     builder/product.cpp
     builder/reshape.cpp
     builder/reverse.cpp

src/ngraph/runtime/cpu/builder/relu.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 "ngraph/runtime/cpu/kernel/relu.hpp"
+#include "ngraph/op/relu.hpp"
+#include "ngraph/runtime/cpu/cpu_builder.hpp"
+#include "ngraph/runtime/cpu/mkldnn_invoke.hpp"
+#include "ngraph/runtime/cpu/mkldnn_utils.hpp"
+
+using namespace std;
+using namespace ngraph;
+
+namespace ngraph
+{
+    namespace runtime
+    {
+        namespace cpu
+        {
+            template <>
+            void Builder::BUILDER_DECL(ngraph::op::ReluBackprop)
+            {
+                auto& functors = external_function->get_functors();
+                auto& tensor_data = external_function->get_tensor_data();
+
+                auto& arg_fwd_tensor = tensor_data[args[0].get_name()];
+                auto& delta_tensor = tensor_data[args[1].get_name()];
+                auto& out_tensor = tensor_data[out[0].get_name()];
+                size_t count = out[0].get_size();
+
+                if (runtime::cpu::mkldnn_utils::use_mkldnn_kernel(node))
+                {
+                    auto& mkldnn_emitter = external_function->get_mkldnn_emitter();
+                    auto input_desc = mkldnn_emitter->build_memory_descriptor(
+                        args[0], runtime::cpu::mkldnn_utils::get_input_mkldnn_format(node, 0));
+                    auto delta_desc = mkldnn_emitter->build_memory_descriptor(
+                        args[1], runtime::cpu::mkldnn_utils::get_input_mkldnn_format(node, 1));
+                    auto result_desc = mkldnn_emitter->build_memory_descriptor(
+                        out[0], runtime::cpu::mkldnn_utils::get_output_mkldnn_format(node, 0));
+
+                    size_t relu_index =
+                        mkldnn_emitter->build_relu_backward(input_desc, delta_desc, result_desc);
+
+                    auto& deps = mkldnn_emitter->get_primitive_deps(relu_index);
+                    auto functor = [&, relu_index](CPURuntimeContext* ctx) {
+                        cpu::mkldnn_utils::set_memory_ptr(ctx, deps[0], arg_fwd_tensor);
+                        cpu::mkldnn_utils::set_memory_ptr(ctx, deps[1], delta_tensor);
+                        cpu::mkldnn_utils::set_memory_ptr(ctx, deps[2], out_tensor);
+                        cpu::mkldnn_utils::mkldnn_invoke_primitive(ctx, relu_index);
+                    };
+                    functors.emplace_back(functor);
+                }
+                else
+                {
+                    std::function<decltype(runtime::cpu::kernel::relu_backprop<float>)> kernel;
+
+                    SELECT_KERNEL(
+                        kernel, out[0].get_element_type(), runtime::cpu::kernel::relu_backprop);
+
+                    auto functor = [&, kernel, count](CPURuntimeContext* ctx) {
+                        kernel(arg_fwd_tensor, delta_tensor, out_tensor, count);
+                    };
+                    functors.emplace_back(functor);
+                }
+            }
+            REGISTER_OP_BUILDER(ReluBackprop);
+        }
+    }
+}

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

@@ -20,6 +20,7 @@
 #include <unsupported/Eigen/CXX11/Tensor>
 
 #include "ngraph/runtime/cpu/kernel/eigen_thread_pool.hpp"
+#include "ngraph/runtime/reference/relu.hpp"
 
 namespace ngraph
 {
@@ -43,6 +44,15 @@ namespace ngraph
 
                     out.device(eigen::global_thread_pool_device) = in0.cwiseMax(ElementType(0));
                 }
+
+                template <typename ElementType>
+                void relu_backprop(void* arg, void* delta_arg, void* out, size_t count)
+                {
+                    reference::relu_backprop<ElementType>(static_cast<ElementType*>(arg),
+                                                          static_cast<ElementType*>(delta_arg),
+                                                          static_cast<ElementType*>(out),
+                                                          count);
+                }
             }
         }
     }