Optimize 4D Reshape (#836)

* CPU: Optimize 4D "nGraph" Reshapes (shuffle+reshape)

* CPU: Add kernel sources

* CPU: Replace 2D with 3D reshape

* CPU: Fixes

* CPU: Simplify

src/ngraph/CMakeLists.txt

@@ -209,6 +209,7 @@ if (NGRAPH_CPU_ENABLE AND LLVM_INCLUDE_DIR AND
         runtime/cpu/kernel/pad.cpp
         runtime/cpu/kernel/reduce_max.cpp
         runtime/cpu/kernel/reduce_sum.cpp
+        runtime/cpu/kernel/reshape.cpp
         runtime/cpu/op/conv_bias.cpp
         runtime/cpu/op/conv_relu.cpp
         runtime/cpu/op/convert_layout.cpp

src/ngraph/runtime/cpu/cpu_emitter.cpp

@@ -1318,13 +1318,36 @@ namespace ngraph
                     writer << "               );\n";
                 }
 #else
-                kernel::emit_reshape(writer,
-                                     args[0].get_element_type().c_type_string(),
-                                     args[0].get_name(),
-                                     out[0].get_name(),
-                                     args[0].get_shape(),
-                                     out[0].get_shape(),
-                                     reshape->get_input_order());
+                if (args[0].get_element_type() == element::f32 && args[0].get_shape().size() == 3 &&
+                    out[0].get_shape().size() == 3)
+                {
+                    writer << "cpu::kernel::reshape_3d_3d_float32(" << args[0].get_name() << ", "
+                           << out[0].get_name() << ", "
+                           << "{" << join(args[0].get_shape()) << "}, "
+                           << "{" << join(reshape->get_input_order()) << "}, "
+                           << "{" << join(out[0].get_shape()) << "}"
+                           << ");\n";
+                }
+                else if (args[0].get_element_type() == element::f32 &&
+                         args[0].get_shape().size() == 4 && out[0].get_shape().size() == 4)
+                {
+                    writer << "cpu::kernel::reshape_4d_4d_float32(" << args[0].get_name() << ", "
+                           << out[0].get_name() << ", "
+                           << "{" << join(args[0].get_shape()) << "}, "
+                           << "{" << join(reshape->get_input_order()) << "}, "
+                           << "{" << join(out[0].get_shape()) << "}"
+                           << ");\n";
+                }
+                else
+                {
+                    kernel::emit_reshape(writer,
+                                         args[0].get_element_type().c_type_string(),
+                                         args[0].get_name(),
+                                         out[0].get_name(),
+                                         args[0].get_shape(),
+                                         out[0].get_shape(),
+                                         reshape->get_input_order());
+                }
 #endif
                 writer.block_end();
             }

src/ngraph/runtime/cpu/cpu_kernels.hpp

@@ -109,6 +109,7 @@ namespace ngraph
 {
     class Shape;
     class AxisSet;
+    class AxisVector;
 
     namespace runtime
     {
@@ -150,6 +151,18 @@ namespace ngraph
                                                const Shape& input_shape,
                                                const Shape& output_shape,
                                                const AxisSet& reduction_axes);
+
+                void reshape_3d_3d_float32(float* input,
+                                           float* output,
+                                           const Shape& input_shape,
+                                           const AxisVector& input_axis_order,
+                                           const Shape& output_shape);
+
+                void reshape_4d_4d_float32(float* input,
+                                           float* output,
+                                           const Shape& input_shape,
+                                           const AxisVector& input_axis_order,
+                                           const Shape& output_shape);
             }
         }
     }

src/ngraph/runtime/cpu/kernel/reshape.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 "reshape.hpp"
+
+namespace ngraph
+{
+    namespace runtime
+    {
+        namespace cpu
+        {
+            namespace kernel
+            {
+                void reshape_3d_3d_float32(float* input,
+                                           float* output,
+                                           const Shape& input_shape,
+                                           const AxisVector& input_axis_order,
+                                           const Shape& output_shape)
+                {
+                    reshape<float, 3, 3>(
+                        input, output, input_shape, input_axis_order, output_shape);
+                }
+
+                void reshape_4d_4d_float32(float* input,
+                                           float* output,
+                                           const Shape& input_shape,
+                                           const AxisVector& input_axis_order,
+                                           const Shape& output_shape)
+                {
+                    reshape<float, 4, 4>(
+                        input, output, input_shape, input_axis_order, output_shape);
+                }
+            }
+        }
+    }
+}

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

+/*******************************************************************************
+* 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.
+*******************************************************************************/
+
+#pragma once
+
+#define EIGEN_USE_THREADS
+#include <unsupported/Eigen/CXX11/Tensor>
+
+#include "ngraph/axis_vector.hpp"
+#include "ngraph/runtime/cpu/kernel/eigen_thread_pool.hpp"
+#include "ngraph/shape.hpp"
+
+namespace ngraph
+{
+    namespace runtime
+    {
+        namespace cpu
+        {
+            namespace kernel
+            {
+                template <typename ElementType, unsigned int InRank, unsigned int OutRank>
+                void reshape(ElementType* input,
+                             ElementType* output,
+                             const Shape& input_shape,
+                             const AxisVector& input_axis_order,
+                             const Shape& output_shape)
+                {
+                    Eigen::array<Eigen::Index, OutRank> out_dims;
+                    Eigen::array<Eigen::Index, InRank> in_dims;
+                    Eigen::array<Eigen::Index, InRank> axis_order;
+
+                    for (int i = 0; i < OutRank; i++)
+                    {
+                        out_dims[i] = output_shape[i];
+                    }
+
+                    for (int i = 0; i < InRank; i++)
+                    {
+                        in_dims[i] = input_shape[i];
+                        axis_order[i] = input_axis_order[i];
+                    }
+
+                    Eigen::TensorMap<Eigen::Tensor<ElementType, OutRank, Eigen::RowMajor>> out(
+                        output, out_dims);
+                    Eigen::TensorMap<Eigen::Tensor<ElementType, InRank, Eigen::RowMajor>> in(
+                        input, in_dims);
+
+                    out.device(eigen::global_thread_pool_device) =
+                        in.shuffle(axis_order).reshape(out_dims);
+                }
+            }
+        }
+    }
+}