WIP

src/ngraph/runtime/cpu/cpu_backend.cpp

@@ -15,7 +15,7 @@
 #include "ngraph/runtime/cpu/cpu_backend.hpp"
 #include "ngraph/log.hpp"
 #include "ngraph/runtime/external_function.hpp"
-#include "ngraph/runtime/host_tensor_view.hpp"
+#include "ngraph/runtime/cpu/cpu_tensor_view.hpp"
 
 using namespace ngraph;
 using namespace std;
@@ -30,6 +30,6 @@ std::shared_ptr<ngraph::runtime::TensorView>
     runtime::cpu::CPU_Backend::make_primary_tensor_view(const ngraph::element::Type& element_type,
                                                         const Shape& shape)
 {
-    auto rc = make_shared<runtime::HostTensorView>(element_type, shape);
+    auto rc = make_shared<runtime::cpu::CPUTensorView>(element_type, shape);
     return dynamic_pointer_cast<runtime::TensorView>(rc);
 }

src/ngraph/runtime/cpu/cpu_call_frame.cpp

@@ -16,7 +16,7 @@
 
 #include "ngraph/runtime/cpu/cpu_call_frame.hpp"
 #include "ngraph/runtime/cpu/cpu_external_function.hpp"
-#include "ngraph/runtime/host_tensor_view.hpp"
+#include "ngraph/runtime/cpu/cpu_tensor_view.hpp"
 
 using namespace std;
 using namespace ngraph;
@@ -36,14 +36,14 @@ void runtime::cpu::CPU_CallFrame::tensor_call(
     vector<void*> outputs;
     for (size_t i = 0; i < input_tvs.size(); i++)
     {
-        shared_ptr<runtime::HostTensorView> tv =
-            static_pointer_cast<runtime::HostTensorView>(input_tvs[i]);
+        shared_ptr<runtime::cpu::CPUTensorView> tv =
+            static_pointer_cast<runtime::cpu::CPUTensorView>(input_tvs[i]);
         inputs.push_back(tv->get_data_ptr());
     }
     for (size_t i = 0; i < output_tvs.size(); i++)
     {
-        shared_ptr<runtime::HostTensorView> tv =
-            static_pointer_cast<runtime::HostTensorView>(output_tvs[i]);
+        shared_ptr<runtime::cpu::CPUTensorView> tv =
+            static_pointer_cast<runtime::cpu::CPUTensorView>(output_tvs[i]);
         outputs.push_back(tv->get_data_ptr());
     }
 

src/ngraph/runtime/cpu/cpu_tensor_view.cpp

@@ -15,13 +15,20 @@
 #include <cstring>
 #include <memory>
 
-#include "cpu_tensor_view.hpp"
+#include "ngraph/except.hpp"
+#include "ngraph/shape.hpp"
 #include "ngraph/descriptor/layout/tensor_view_layout.hpp"
 #include "ngraph/descriptor/primary_tensor_view.hpp"
+#include "cpu_tensor_view.hpp"
 
 using namespace ngraph;
 using namespace std;
 
+// TODO(jmenon): Refactor all the alignment specifications into
+// a single place and allow lower or no alignment when possible
+
+const size_t runtime::cpu::CPUTensorView::BufferAlignment = 64;
+
 runtime::cpu::CPUTensorView::CPUTensorView(const ngraph::element::Type& element_type,
                                            const Shape& shape,
                                            const string& name)
@@ -30,6 +37,19 @@ runtime::cpu::CPUTensorView::CPUTensorView(const ngraph::element::Type& element_
     , buffer(nullptr)
     , aligned_buffer(nullptr)
 {
+    buffer_size = shape_size(shape) * element_type.size();
+    if (buffer_size)
+    {
+        size_t allocation_size = buffer_size + BufferAlignment;
+        auto ptr = malloc(allocation_size);
+        if (!ptr)
+        {
+            throw ngraph_error("Error allocating CPU Tensor View memory");
+        }
+        buffer = static_cast<char*>(ptr);
+        std::align(BufferAlignment, buffer_size, ptr, allocation_size);
+        aligned_buffer = static_cast<char*>(ptr);
+    }
 }
 
 runtime::cpu::CPUTensorView::~CPUTensorView()
@@ -69,10 +89,5 @@ void runtime::cpu::CPUTensorView::read(void* target, size_t tensor_offset, size_
 
 size_t runtime::cpu::CPUTensorView::get_size() const
 {
-    return get_tensor_view_layout()->get_size();
-}
-
-const element::Type& runtime::cpu::CPUTensorView::get_element_type() const
-{
-    return get_tensor_view_layout()->get_element_type();
+    return get_element_count();
 }

src/ngraph/runtime/cpu/cpu_tensor_view.hpp

@@ -52,6 +52,8 @@ namespace ngraph
                 void read(void* p, size_t tensor_offset, size_t n) const override;
 
             private:
+                static const size_t BufferAlignment;
+
                 char* buffer;
                 char* aligned_buffer;
                 size_t buffer_size;

src/ngraph/runtime/tensor_view.cpp

@@ -47,6 +47,11 @@ const Strides& runtime::TensorView::get_strides() const
     return m_descriptor->get_tensor_view_layout()->get_strides();
 }
 
+const element::Type& runtime::TensorView::get_element_type() const
+{
+    return m_descriptor->get_tensor_view_type()->get_element_type();
+}
+
 shared_ptr<descriptor::layout::TensorViewLayout> runtime::TensorView::get_tensor_view_layout() const
 {
     return m_descriptor->get_tensor_view_layout();

src/ngraph/runtime/tensor_view.hpp

@@ -20,6 +20,7 @@
 #include "ngraph/descriptor/tensor_view.hpp"
 #include "ngraph/shape.hpp"
 #include "ngraph/util.hpp"
+#include "ngraph/types/element_type.hpp"
 
 namespace ngraph
 {
@@ -52,6 +53,7 @@ namespace ngraph
 
             const ngraph::Shape& get_shape() const;
             const ngraph::Strides& get_strides() const;
+            const ngraph::element::Type& get_element_type() const;
             size_t get_element_count() const;
             const ngraph::descriptor::Tensor& get_tensor() const;
 

test/util/test_tools.hpp

@@ -40,7 +40,7 @@ void copy_data(std::shared_ptr<ngraph::runtime::TensorView> tv, const std::vecto
 template <typename T>
 std::vector<T> read_vector(std::shared_ptr<ngraph::runtime::TensorView> tv)
 {
-    if (ngraph::element::from<T>() != tv->get_tensor_view_layout()->get_element_type())
+    if (ngraph::element::from<T>() != tv->get_element_type())
     {
         throw std::invalid_argument("read_vector type must match TensorView type");
     }