restore fetch utility for opencl (#1770)

src/framework/executor.cpp

@@ -460,6 +460,20 @@ std::shared_ptr<LoDTensor> Executor<Device, T>::GetOutput(
   }
 }
 
+#ifdef PADDLE_MOBILE_CL
+template <typename Device, typename T>
+const CLImage *Executor<Device, T>::GetOutputImage(
+    const std::string &var_name) {
+  auto var = program_.scope->FindVar(var_name);
+  if (var->IsInitialized() && var->template IsType<framework::CLImage>()) {
+    const CLImage *cl_image = var->template Get<framework::CLImage>();
+    return cl_image;
+  } else {
+    return nullptr;
+  }
+}
+#endif
+
 template <typename Device, typename T>
 PMStatus Executor<Device, T>::Predict() {
 #if _OPENMP

src/framework/executor.h

@@ -53,6 +53,9 @@ class Executor {
   void SetInput(const LoDTensor &input, const std::string &var_name);
 
   std::shared_ptr<LoDTensor> GetOutput(const std::string &var_name);
+#ifdef PADDLE_MOBILE_CL
+  const CLImage *GetOutputImage(const std::string &var_name);
+#endif
 
   void FeedTensorData(const std::vector<framework::Tensor> &v);
   void GetTensorResults(std::vector<framework::Tensor *> *v);

src/io/paddle_mobile.cpp

@@ -170,6 +170,14 @@ LoDTensorPtr PaddleMobile<Device, T>::Fetch(const std::string &var_name) {
   return executor_->GetOutput(var_name);
 }
 
+#ifdef PADDLE_MOBILE_CL
+template <typename Device, typename T>
+const framework::CLImage *PaddleMobile<Device, T>::FetchImage(
+    const std::string &var_name) {
+  return executor_->GetOutputImage(var_name);
+}
+#endif
+
 template <typename Device, typename T>
 void PaddleMobile<Device, T>::Clear() {
   executor_ = nullptr;

src/io/paddle_mobile.h

@@ -74,6 +74,9 @@ class PaddleMobile {
 
   typedef std::shared_ptr<framework::LoDTensor> LoDTensorPtr;
   LoDTensorPtr Fetch(const std::string &var_name);
+#ifdef PADDLE_MOBILE_CL
+  const framework::CLImage *FetchImage(const std::string &var_name);
+#endif
 
   LoDTensorPtr Fetch() { return Fetch("fetch"); }
 

test/net/test_net.cpp

@@ -167,6 +167,44 @@ void test(int argc, char *argv[]) {
       paddle_mobile.Feed(var_names[0], input_tensor);
       paddle_mobile.Predict();
     }
+#ifdef PADDLE_MOBILE_CL
+    for (auto var_name : var_names) {
+      auto cl_image = paddle_mobile.FetchImage(var_name);
+      auto len = cl_image->numel();
+      if (len == 0) {
+        continue;
+      }
+      int width = cl_image->ImageDims()[0];
+      int height = cl_image->ImageDims()[1];
+      paddle_mobile::framework::half_t *image_data =
+          new paddle_mobile::framework::half_t[height * width * 4];
+      cl_int err;
+      cl_mem image = cl_image->GetCLImage();
+      size_t origin[3] = {0, 0, 0};
+      size_t region[3] = {width, height, 1};
+      err = clEnqueueReadImage(cl_image->CommandQueue(), image, CL_TRUE, origin,
+                               region, 0, 0, image_data, 0, NULL, NULL);
+      CL_CHECK_ERRORS(err);
+      float *tensor_data = new float[cl_image->numel()];
+      auto converter = cl_image->Converter();
+      converter->ImageToNCHW(image_data, tensor_data, cl_image->ImageDims(),
+                             cl_image->dims());
+
+      auto data = tensor_data;
+      std::string sample = "";
+      if (!is_sample_step) {
+        sample_step = len / sample_num;
+      }
+      if (sample_step <= 0) {
+        sample_step = 1;
+      }
+      for (int i = 0; i < len; i += sample_step) {
+        sample += " " + std::to_string(data[i]);
+      }
+      std::cout << "auto-test"
+                << " var " << var_name << sample << std::endl;
+    }
+#else
     for (auto var_name : var_names) {
       auto out = paddle_mobile.Fetch(var_name);
       auto len = out->numel();
@@ -206,6 +244,7 @@ void test(int argc, char *argv[]) {
                   << " var " << var_name << sample << std::endl;
       }
     }
+#endif
     std::cout << std::endl;
   }
 }

test/test_include.h

@@ -33,3 +33,7 @@ limitations under the License. */
 #include "framework/tensor.h"
 #include "framework/variable.h"
 #include "io/paddle_mobile.h"
+
+#ifdef PADDLE_MOBILE_CL
+#include "framework/cl/cl_image.h"
+#endif