Rewrite the way constants are emitted in the CPU backend (#332)

* wip

* constants as globals

* const emitter rewrite

src/ngraph/runtime/cpu/cpu_emitter.cpp

@@ -573,19 +573,6 @@ void runtime::cpu::CPU_Emitter::EmitConstant(const ngraph::Node* n,
                                              const vector<runtime::cpu::TensorViewWrapper>& args,
                                              const vector<runtime::cpu::TensorViewWrapper>& out)
 {
-    auto c = static_cast<const op::Constant*>(n);
-    auto c_value_strings = c->get_value_strings();
-    auto type = out[0].get_type();
-
-    m_out << "{   // " << n->get_name() << " EmitConstant\n";
-    m_out.indent++;
-    for (size_t i = 0; i < c_value_strings.size(); i++)
-    {
-        m_out << out[0].get_name() << "[" << i << "] = static_cast<" << type << ">("
-              << c_value_strings[i] << ");\n";
-    }
-    m_out.indent--;
-    m_out << "}\n";
 }
 
 void runtime::cpu::CPU_Emitter::EmitReshape(const ngraph::Node* n,

src/ngraph/runtime/cpu/cpu_external_function.cpp

@@ -292,6 +292,28 @@ using namespace ngraph::runtime;
         writer << "\n";
     }
 
+    writer << "// Declare all constants\n";
+    for (shared_ptr<Function> current_function : pass_manager.get_state().get_functions())
+    {
+        for (shared_ptr<Node> node : current_function->get_ordered_ops())
+        {
+            const op::Constant* c = dynamic_cast<op::Constant*>(node.get());
+            if (c)
+            {
+                shared_ptr<descriptor::TensorView> tv = node->get_outputs()[0].get_tensor_view();
+                auto c_value_strings = c->get_value_strings();
+                writer << tv->get_tensor().get_element_type().c_type_string() << " "
+                       << tv->get_tensor().get_name() << "[" << c_value_strings.size() << "] =\n";
+                writer << "{\n";
+                for (size_t i = 0; i < c_value_strings.size(); i++)
+                {
+                    writer << "    " << c_value_strings[i] << ",\n";
+                }
+                writer << "};\n\n";
+            }
+        }
+    }
+
     writer << "// Declare all functions\n";
     for (shared_ptr<Function> f : pass_manager.get_state().get_functions())
     {
@@ -301,6 +323,22 @@ using namespace ngraph::runtime;
 
     for (shared_ptr<Function> current_function : pass_manager.get_state().get_functions())
     {
+        set<string> output_names;
+        for (const descriptor::Output* output : current_function->get_outputs())
+        {
+            shared_ptr<descriptor::TensorView> tv = output->get_tensor_view();
+            output_names.insert(tv->get_tensor().get_name());
+        }
+        set<descriptor::TensorView*> constants;
+        for (shared_ptr<Node> node : current_function->get_ordered_ops())
+        {
+            if (dynamic_cast<op::Constant*>(node.get()))
+            {
+                shared_ptr<descriptor::TensorView> tv = node->get_outputs()[0].get_tensor_view();
+                constants.insert(tv.get());
+            }
+        }
+
         writer << "extern \"C\" void " << current_function->get_name();
         writer << "(void** inputs, void** outputs)\n";
         writer << "{\n";
@@ -360,7 +398,6 @@ using namespace ngraph::runtime;
         writer << "\n";
 
         writer << "// Define outputs\n";
-
         // create alias list
         size_t output_index = 0;
         unordered_map<descriptor::TensorView*, vector<size_t>> output_alias_map;
@@ -378,7 +415,6 @@ using namespace ngraph::runtime;
         }
 
         output_index = 0;
-        set<string> output_names;
         for (const descriptor::Output* output : current_function->get_outputs())
         {
             shared_ptr<descriptor::TensorView> tv = output->get_tensor_view();
@@ -401,28 +437,20 @@ using namespace ngraph::runtime;
             }
             if (!parameter_as_output && !contains(aliases, output_index))
             {
-                string type = et.c_type_string();
-                writer << type << "* " << tv->get_tensor().get_name() << " = static_cast<" << type
-                       << "*>(outputs[" << output_index << "]);\n";
-            }
-            output_names.insert(tv->get_tensor().get_name());
-            output_index++;
-        }
-        writer << "\n";
-
-        writer << "// Define constants\n";
-        for (shared_ptr<Node> node : current_function->get_ordered_ops())
-        {
-            if (dynamic_cast<op::Constant*>(node.get()))
-            {
-                shared_ptr<descriptor::TensorView> tv = node->get_outputs()[0].get_tensor_view();
-                if (!contains(output_names, tv->get_tensor().get_name()))
+                if (contains(constants, tv.get()))
                 {
-                    writer << tv->get_tensor().get_element_type().c_type_string() << " "
-                           << tv->get_tensor().get_name() << "[" << tv->get_tensor().size()
-                           << "];\n";
+                    writer << "memcpy(outputs[" << output_index << "], "
+                           << tv->get_tensor().get_name() << ", " << tv->get_tensor().size()
+                           << ");\n";
+                }
+                else
+                {
+                    string type = et.c_type_string();
+                    writer << type << "* " << tv->get_tensor().get_name() << " = static_cast<"
+                           << type << "*>(outputs[" << output_index << "]);\n";
                 }
             }
+            output_index++;
         }
         writer << "\n";