refactor(tensorrt_common)!: fix namespace, directory structure & move…

… to perception namespace (#9099) * refactor(tensorrt_common)!: fix namespace, directory structure & move to perception namespace Signed-off-by: amadeuszsz <[email protected]> * refactor(tensorrt_common): directory structure Signed-off-by: amadeuszsz <[email protected]> * style(pre-commit): autofix * fix(tensorrt_common): correct package name for logging Signed-off-by: amadeuszsz <[email protected]> --------- Signed-off-by: amadeuszsz <[email protected]> Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> Co-authored-by: Kenzo Lobos Tsunekawa <[email protected]>
autowarefoundation · Oct 21, 2024 · 35d5fbf · 35d5fbf
1 parent 20ebe9d
commit 35d5fbf
Show file tree

Hide file tree

Showing 40 changed files with 179 additions and 160 deletions.
diff --git a/.github/CODEOWNERS b/.github/CODEOWNERS
@@ -28,7 +28,6 @@ common/goal_distance_calculator/** [email protected]
 common/osqp_interface/** [email protected] [email protected] [email protected] [email protected]
 common/polar_grid/** [email protected]
 common/qp_interface/** [email protected] [email protected] [email protected] [email protected]
-common/tensorrt_common/** [email protected] [email protected] [email protected] [email protected]
 common/tier4_adapi_rviz_plugin/** [email protected] [email protected] [email protected]
 common/tier4_api_utils/** [email protected]
 common/tier4_camera_view_rviz_plugin/** [email protected] [email protected]
@@ -132,6 +131,7 @@ perception/autoware_raindrop_cluster_filter/** [email protected] yoshi.ri@tier
 perception/autoware_shape_estimation/** [email protected] [email protected] [email protected]
 perception/autoware_simple_object_merger/** [email protected] [email protected] [email protected]
 perception/autoware_tensorrt_classifier/** [email protected] [email protected]
+perception/autoware_tensorrt_common/** [email protected] [email protected] [email protected] [email protected]
 perception/autoware_tensorrt_yolox/** [email protected] [email protected] [email protected]
 perception/autoware_tracking_object_merger/** [email protected] [email protected] [email protected]
 perception/autoware_traffic_light_arbiter/** [email protected] [email protected]

diff --git a/.github/labeler.yaml b/.github/labeler.yaml
@@ -42,5 +42,3 @@
 "tag:require-cuda-build-and-test":
   - perception/**/*
   - sensing/**/*
-  - common/cuda_utils/**/*
-  - common/tensorrt_common/**/*
diff --git a/common/.pages b/common/.pages
@@ -9,7 +9,6 @@ nav:
       - 'comparisons': common/autoware_auto_common/design/comparisons
     - 'autoware_grid_map_utils': common/autoware_grid_map_utils
     - 'autoware_point_types': common/autoware_point_types
-    - 'Cuda Utils': common/cuda_utils
     - 'Geography Utils': common/geography_utils
     - 'Global Parameter Loader': common/global_parameter_loader/Readme
     - 'Glog Component': common/glog_component
@@ -24,7 +23,6 @@ nav:
     - 'Signal Processing':
       - 'Introduction': common/signal_processing
       - 'Butterworth Filter': common/signal_processing/documentation/ButterworthFilter
-    - 'TensorRT Common': common/tensorrt_common
     - 'autoware_universe_utils': common/autoware_universe_utils
     - 'traffic_light_utils': common/traffic_light_utils
     - 'TVM Utility':

diff --git a/perception/autoware_bytetrack/package.xml b/perception/autoware_bytetrack/package.xml
@@ -17,6 +17,7 @@
 
   <depend>autoware_kalman_filter</depend>
   <depend>autoware_perception_msgs</depend>
+  <depend>autoware_tensorrt_common</depend>
   <depend>cuda_utils</depend>
   <depend>cv_bridge</depend>
   <depend>eigen</depend>
@@ -26,7 +27,6 @@
   <depend>rclcpp</depend>
   <depend>rclcpp_components</depend>
   <depend>sensor_msgs</depend>
-  <depend>tensorrt_common</depend>
   <depend>tier4_perception_msgs</depend>
   <depend>yaml-cpp</depend>
 

diff --git a/perception/autoware_lidar_apollo_instance_segmentation/CMakeLists.txt b/perception/autoware_lidar_apollo_instance_segmentation/CMakeLists.txt
@@ -12,9 +12,9 @@ find_package(autoware_universe_utils REQUIRED)
 find_package(tier4_debug_msgs REQUIRED)
 find_package(tier4_perception_msgs REQUIRED)
 
-find_package(tensorrt_common)
-if(NOT ${tensorrt_common_FOUND})
-  message(WARNING "The tensorrt_common package is not found. Please check its dependencies.")
+find_package(autoware_tensorrt_common)
+if(NOT ${autoware_tensorrt_common_FOUND})
+  message(WARNING "The autoware_tensorrt_common package is not found. Please check its dependencies.")
   # to avoid launch file missing without a gpu
   ament_package()
   install(
@@ -48,8 +48,8 @@ target_link_libraries(${PROJECT_NAME}
   pcl_common
   rclcpp::rclcpp
   rclcpp_components::component
-  tensorrt_common::tensorrt_common
   tf2_eigen::tf2_eigen
+  ${autoware_tensorrt_common_TARGETS}
   ${tier4_debug_msgs_TARGETS}
   ${tier4_perception_msgs_TARGETS}
 )

diff --git a/...lo_instance_segmentation/include/autoware/lidar_apollo_instance_segmentation/detector.hpp b/...lo_instance_segmentation/include/autoware/lidar_apollo_instance_segmentation/detector.hpp
@@ -19,10 +19,10 @@
 #include "autoware/lidar_apollo_instance_segmentation/feature_generator.hpp"
 #include "autoware/lidar_apollo_instance_segmentation/node.hpp"
 
+#include <autoware/tensorrt_common/tensorrt_common.hpp>
 #include <autoware/universe_utils/transform/transforms.hpp>
 #include <cuda_utils/cuda_unique_ptr.hpp>
 #include <cuda_utils/stream_unique_ptr.hpp>
-#include <tensorrt_common/tensorrt_common.hpp>
 #include <tf2_eigen/tf2_eigen.hpp>
 
 #include <tf2_ros/buffer_interface.h>
@@ -54,7 +54,7 @@ class LidarApolloInstanceSegmentation : public LidarInstanceSegmentationInterfac
     const sensor_msgs::msg::PointCloud2 & input, sensor_msgs::msg::PointCloud2 & transformed_cloud,
     float z_offset);
 
-  std::unique_ptr<tensorrt_common::TrtCommon> trt_common_;
+  std::unique_ptr<autoware::tensorrt_common::TrtCommon> trt_common_;
 
   rclcpp::Node * node_;
   std::shared_ptr<Cluster2D> cluster2d_;

diff --git a/perception/autoware_lidar_apollo_instance_segmentation/package.xml b/perception/autoware_lidar_apollo_instance_segmentation/package.xml
@@ -14,13 +14,13 @@
   <buildtool_depend>ament_cmake</buildtool_depend>
 
   <depend>autoware_perception_msgs</depend>
+  <depend>autoware_tensorrt_common</depend>
   <depend>autoware_universe_utils</depend>
   <depend>cuda_utils</depend>
   <depend>libpcl-all-dev</depend>
   <depend>pcl_conversions</depend>
   <depend>rclcpp</depend>
   <depend>rclcpp_components</depend>
-  <depend>tensorrt_common</depend>
   <depend>tf2_eigen</depend>
   <depend>tier4_debug_msgs</depend>
   <depend>tier4_perception_msgs</depend>

diff --git a/perception/autoware_lidar_apollo_instance_segmentation/src/detector.cpp b/perception/autoware_lidar_apollo_instance_segmentation/src/detector.cpp
@@ -47,8 +47,8 @@ LidarApolloInstanceSegmentation::LidarApolloInstanceSegmentation(rclcpp::Node *
   z_offset_ = node_->declare_parameter<float>("z_offset", -2.0);
   const auto precision = node_->declare_parameter("precision", "fp32");
 
-  trt_common_ = std::make_unique<tensorrt_common::TrtCommon>(
-    onnx_file, precision, nullptr, tensorrt_common::BatchConfig{1, 1, 1}, 1 << 30);
+  trt_common_ = std::make_unique<autoware::tensorrt_common::TrtCommon>(
+    onnx_file, precision, nullptr, autoware::tensorrt_common::BatchConfig{1, 1, 1}, 1 << 30);
   trt_common_->setup();
 
   if (!trt_common_->isInitialized()) {

diff --git a/...utoware_lidar_centerpoint/include/autoware/lidar_centerpoint/network/tensorrt_wrapper.hpp b/...utoware_lidar_centerpoint/include/autoware/lidar_centerpoint/network/tensorrt_wrapper.hpp
@@ -17,7 +17,7 @@
 
 #include "NvInfer.h"
 #include "autoware/lidar_centerpoint/centerpoint_config.hpp"
-#include "tensorrt_common/tensorrt_common.hpp"
+#include "autoware/tensorrt_common/tensorrt_common.hpp"
 
 #include <iostream>
 #include <memory>
@@ -36,15 +36,15 @@ class TensorRTWrapper
   bool init(
     const std::string & onnx_path, const std::string & engine_path, const std::string & precision);
 
-  tensorrt_common::TrtUniquePtr<nvinfer1::IExecutionContext> context_{nullptr};
+  autoware::tensorrt_common::TrtUniquePtr<nvinfer1::IExecutionContext> context_{nullptr};
 
 protected:
   virtual bool setProfile(
     nvinfer1::IBuilder & builder, nvinfer1::INetworkDefinition & network,
     nvinfer1::IBuilderConfig & config) = 0;
 
   CenterPointConfig config_;
-  tensorrt_common::Logger logger_;
+  autoware::tensorrt_common::Logger logger_;
 
 private:
   bool parseONNX(
@@ -57,9 +57,9 @@ class TensorRTWrapper
 
   bool createContext();
 
-  tensorrt_common::TrtUniquePtr<nvinfer1::IRuntime> runtime_{nullptr};
-  tensorrt_common::TrtUniquePtr<nvinfer1::IHostMemory> plan_{nullptr};
-  tensorrt_common::TrtUniquePtr<nvinfer1::ICudaEngine> engine_{nullptr};
+  autoware::tensorrt_common::TrtUniquePtr<nvinfer1::IRuntime> runtime_{nullptr};
+  autoware::tensorrt_common::TrtUniquePtr<nvinfer1::IHostMemory> plan_{nullptr};
+  autoware::tensorrt_common::TrtUniquePtr<nvinfer1::ICudaEngine> engine_{nullptr};
 };
 
 }  // namespace autoware::lidar_centerpoint

diff --git a/perception/autoware_lidar_centerpoint/lib/network/network_trt.cpp b/perception/autoware_lidar_centerpoint/lib/network/network_trt.cpp
@@ -63,7 +63,7 @@ bool HeadTRT::setProfile(
     if (
       out_name == std::string("heatmap") &&
       network.getOutput(ci)->getDimensions().d[1] != static_cast<int32_t>(out_channel_sizes_[ci])) {
-      tensorrt_common::LOG_ERROR(logger_)
+      autoware::tensorrt_common::LOG_ERROR(logger_)
         << "Expected and actual number of classes do not match" << std::endl;
       return false;
     }

diff --git a/perception/autoware_lidar_centerpoint/lib/network/tensorrt_wrapper.cpp b/perception/autoware_lidar_centerpoint/lib/network/tensorrt_wrapper.cpp
@@ -37,10 +37,10 @@ TensorRTWrapper::~TensorRTWrapper()
 bool TensorRTWrapper::init(
   const std::string & onnx_path, const std::string & engine_path, const std::string & precision)
 {
-  runtime_ =
-    tensorrt_common::TrtUniquePtr<nvinfer1::IRuntime>(nvinfer1::createInferRuntime(logger_));
+  runtime_ = autoware::tensorrt_common::TrtUniquePtr<nvinfer1::IRuntime>(
+    nvinfer1::createInferRuntime(logger_));
   if (!runtime_) {
-    tensorrt_common::LOG_ERROR(logger_) << "Failed to create runtime" << std::endl;
+    autoware::tensorrt_common::LOG_ERROR(logger_) << "Failed to create runtime" << std::endl;
     return false;
   }
 
@@ -63,15 +63,15 @@ bool TensorRTWrapper::init(
 bool TensorRTWrapper::createContext()
 {
   if (!engine_) {
-    tensorrt_common::LOG_ERROR(logger_)
+    autoware::tensorrt_common::LOG_ERROR(logger_)
       << "Failed to create context: Engine was not created" << std::endl;
     return false;
   }
 
-  context_ =
-    tensorrt_common::TrtUniquePtr<nvinfer1::IExecutionContext>(engine_->createExecutionContext());
+  context_ = autoware::tensorrt_common::TrtUniquePtr<nvinfer1::IExecutionContext>(
+    engine_->createExecutionContext());
   if (!context_) {
-    tensorrt_common::LOG_ERROR(logger_) << "Failed to create context" << std::endl;
+    autoware::tensorrt_common::LOG_ERROR(logger_) << "Failed to create context" << std::endl;
     return false;
   }
 
@@ -82,17 +82,17 @@ bool TensorRTWrapper::parseONNX(
   const std::string & onnx_path, const std::string & engine_path, const std::string & precision,
   const std::size_t workspace_size)
 {
-  auto builder =
-    tensorrt_common::TrtUniquePtr<nvinfer1::IBuilder>(nvinfer1::createInferBuilder(logger_));
+  auto builder = autoware::tensorrt_common::TrtUniquePtr<nvinfer1::IBuilder>(
+    nvinfer1::createInferBuilder(logger_));
   if (!builder) {
-    tensorrt_common::LOG_ERROR(logger_) << "Failed to create builder" << std::endl;
+    autoware::tensorrt_common::LOG_ERROR(logger_) << "Failed to create builder" << std::endl;
     return false;
   }
 
-  auto config =
-    tensorrt_common::TrtUniquePtr<nvinfer1::IBuilderConfig>(builder->createBuilderConfig());
+  auto config = autoware::tensorrt_common::TrtUniquePtr<nvinfer1::IBuilderConfig>(
+    builder->createBuilderConfig());
   if (!config) {
-    tensorrt_common::LOG_ERROR(logger_) << "Failed to create config" << std::endl;
+    autoware::tensorrt_common::LOG_ERROR(logger_) << "Failed to create config" << std::endl;
     return false;
   }
 #if (NV_TENSORRT_MAJOR * 1000) + (NV_TENSORRT_MINOR * 100) + NV_TENSOR_PATCH >= 8400
@@ -102,42 +102,43 @@ bool TensorRTWrapper::parseONNX(
 #endif
   if (precision == "fp16") {
     if (builder->platformHasFastFp16()) {
-      tensorrt_common::LOG_INFO(logger_) << "Using TensorRT FP16 Inference" << std::endl;
+      autoware::tensorrt_common::LOG_INFO(logger_) << "Using TensorRT FP16 Inference" << std::endl;
       config->setFlag(nvinfer1::BuilderFlag::kFP16);
     } else {
-      tensorrt_common::LOG_INFO(logger_)
+      autoware::tensorrt_common::LOG_INFO(logger_)
         << "TensorRT FP16 Inference isn't supported in this environment" << std::endl;
     }
   }
 
   const auto flag =
     1U << static_cast<uint32_t>(nvinfer1::NetworkDefinitionCreationFlag::kEXPLICIT_BATCH);
-  auto network =
-    tensorrt_common::TrtUniquePtr<nvinfer1::INetworkDefinition>(builder->createNetworkV2(flag));
+  auto network = autoware::tensorrt_common::TrtUniquePtr<nvinfer1::INetworkDefinition>(
+    builder->createNetworkV2(flag));
   if (!network) {
-    tensorrt_common::LOG_ERROR(logger_) << "Failed to create network" << std::endl;
+    autoware::tensorrt_common::LOG_ERROR(logger_) << "Failed to create network" << std::endl;
     return false;
   }
 
-  auto parser = tensorrt_common::TrtUniquePtr<nvonnxparser::IParser>(
+  auto parser = autoware::tensorrt_common::TrtUniquePtr<nvonnxparser::IParser>(
     nvonnxparser::createParser(*network, logger_));
   parser->parseFromFile(onnx_path.c_str(), static_cast<int>(nvinfer1::ILogger::Severity::kERROR));
 
   if (!setProfile(*builder, *network, *config)) {
-    tensorrt_common::LOG_ERROR(logger_) << "Failed to set profile" << std::endl;
+    autoware::tensorrt_common::LOG_ERROR(logger_) << "Failed to set profile" << std::endl;
     return false;
   }
 
-  plan_ = tensorrt_common::TrtUniquePtr<nvinfer1::IHostMemory>(
+  plan_ = autoware::tensorrt_common::TrtUniquePtr<nvinfer1::IHostMemory>(
     builder->buildSerializedNetwork(*network, *config));
   if (!plan_) {
-    tensorrt_common::LOG_ERROR(logger_) << "Failed to create serialized network" << std::endl;
+    autoware::tensorrt_common::LOG_ERROR(logger_)
+      << "Failed to create serialized network" << std::endl;
     return false;
   }
-  engine_ = tensorrt_common::TrtUniquePtr<nvinfer1::ICudaEngine>(
+  engine_ = autoware::tensorrt_common::TrtUniquePtr<nvinfer1::ICudaEngine>(
     runtime_->deserializeCudaEngine(plan_->data(), plan_->size()));
   if (!engine_) {
-    tensorrt_common::LOG_ERROR(logger_) << "Failed to create engine" << std::endl;
+    autoware::tensorrt_common::LOG_ERROR(logger_) << "Failed to create engine" << std::endl;
     return false;
   }
 
@@ -146,7 +147,7 @@ bool TensorRTWrapper::parseONNX(
 
 bool TensorRTWrapper::saveEngine(const std::string & engine_path)
 {
-  tensorrt_common::LOG_INFO(logger_) << "Writing to " << engine_path << std::endl;
+  autoware::tensorrt_common::LOG_INFO(logger_) << "Writing to " << engine_path << std::endl;
   std::ofstream file(engine_path, std::ios::out | std::ios::binary);
   file.write(reinterpret_cast<const char *>(plan_->data()), plan_->size());
   return true;
@@ -158,9 +159,10 @@ bool TensorRTWrapper::loadEngine(const std::string & engine_path)
   std::stringstream engine_buffer;
   engine_buffer << engine_file.rdbuf();
   std::string engine_str = engine_buffer.str();
-  engine_ = tensorrt_common::TrtUniquePtr<nvinfer1::ICudaEngine>(runtime_->deserializeCudaEngine(
-    reinterpret_cast<const void *>(engine_str.data()), engine_str.size()));
-  tensorrt_common::LOG_INFO(logger_) << "Loaded engine from " << engine_path << std::endl;
+  engine_ =
+    autoware::tensorrt_common::TrtUniquePtr<nvinfer1::ICudaEngine>(runtime_->deserializeCudaEngine(
+      reinterpret_cast<const void *>(engine_str.data()), engine_str.size()));
+  autoware::tensorrt_common::LOG_INFO(logger_) << "Loaded engine from " << engine_path << std::endl;
   return true;
 }
 

diff --git a/perception/autoware_lidar_centerpoint/package.xml b/perception/autoware_lidar_centerpoint/package.xml
@@ -14,11 +14,11 @@
 
   <depend>autoware_object_recognition_utils</depend>
   <depend>autoware_perception_msgs</depend>
+  <depend>autoware_tensorrt_common</depend>
   <depend>autoware_universe_utils</depend>
   <depend>pcl_ros</depend>
   <depend>rclcpp</depend>
   <depend>rclcpp_components</depend>
-  <depend>tensorrt_common</depend>
   <depend>tf2_eigen</depend>
   <depend>tf2_geometry_msgs</depend>
   <depend>tf2_sensor_msgs</depend>

diff --git a/...ion/autoware_lidar_transfusion/include/autoware/lidar_transfusion/network/network_trt.hpp b/...ion/autoware_lidar_transfusion/include/autoware/lidar_transfusion/network/network_trt.hpp
@@ -18,7 +18,7 @@
 #include "autoware/lidar_transfusion/transfusion_config.hpp"
 #include "autoware/lidar_transfusion/utils.hpp"
 
-#include <tensorrt_common/tensorrt_common.hpp>
+#include <autoware/tensorrt_common/tensorrt_common.hpp>
 
 #include <NvInfer.h>
 
@@ -57,8 +57,8 @@ class NetworkTRT
     const std::string & onnx_path, const std::string & engine_path, const std::string & precision);
   const char * getTensorName(NetworkIO name);
 
-  tensorrt_common::TrtUniquePtr<nvinfer1::ICudaEngine> engine{nullptr};
-  tensorrt_common::TrtUniquePtr<nvinfer1::IExecutionContext> context{nullptr};
+  autoware::tensorrt_common::TrtUniquePtr<nvinfer1::ICudaEngine> engine{nullptr};
+  autoware::tensorrt_common::TrtUniquePtr<nvinfer1::IExecutionContext> context{nullptr};
 
 private:
   bool parseONNX(
@@ -73,9 +73,9 @@ class NetworkTRT
   bool validateNetworkIO();
   nvinfer1::Dims validateTensorShape(NetworkIO name, const std::vector<int> shape);
 
-  tensorrt_common::TrtUniquePtr<nvinfer1::IRuntime> runtime_{nullptr};
-  tensorrt_common::TrtUniquePtr<nvinfer1::IHostMemory> plan_{nullptr};
-  tensorrt_common::Logger logger_;
+  autoware::tensorrt_common::TrtUniquePtr<nvinfer1::IRuntime> runtime_{nullptr};
+  autoware::tensorrt_common::TrtUniquePtr<nvinfer1::IHostMemory> plan_{nullptr};
+  autoware::tensorrt_common::Logger logger_;
   TransfusionConfig config_;
   std::vector<const char *> tensors_names_;