如何在ROS 功能包中编译CUDA .cu文件

在ROS中编译CUDA .cu文件

关键点

1. find_package( CUDA REQUIRED )
2. set(CUDA_NVCC_FLAGS "${CUDA_NVCC_FLAGS} -DMY_DEF=1")
3. set( CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -DMY_DEF=1" )
4. set( CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -DMY_DEF=1" )
5. add_library() 和 add_executable() 用 cuda_add_library() 和 cuda_add_executable() 替代。（针对低版本的cmake，高版本好像不区分）

成功案例

cu文件

#include 
#include 

/*
 * Host function to initialize vector elements. This function
 * simply initializes each element to equal its index in the
 * vector.
 */

void initWith(float num, float *a, int N)
{
  for(int i = 0; i < N; ++i)
  {
    a[i] = num;
  }
}

/*
 * Device kernel stores into `result` the sum of each
 * same-indexed value of `a` and `b`.
 */

__global__
void addVectorsInto(float *result, float *a, float *b, int N)
{
  int index = threadIdx.x + blockIdx.x * blockDim.x;
  int stride = blockDim.x * gridDim.x;

  for(int i = index; i < N; i += stride)
  {
    result[i] = a[i] + b[i];
  }
}

/*
 * Host function to confirm values in `vector`. This function
 * assumes all values are the same `target` value.
 */

void checkElementsAre(float target, float *vector, int N)
{
  for(int i = 0; i < N; i++)
  {
    if(vector[i] != target)
    {
      printf("FAIL: vector[%d] - %0.0f does not equal %0.0f\n", i, vector[i], target);
      exit(1);
    }
  }
  printf("Success! All values calculated correctly.\n");
}

int main(int argc, char* argv[])
{
  //执行 ros 节点初始化
  ros::init(argc,argv,"hello");
  //创建 ros 节点句柄(非必须)
  ros::NodeHandle n;
  //控制台输出 hello world
  ROS_INFO("hello world!");

  const long N = 2<<24;
  size_t size = N * sizeof(float);

  float *a;
  float *b;
  float *c;

  cudaMallocManaged(&a, size);
  cudaMallocManaged(&b, size);
  cudaMallocManaged(&c, size);

  initWith(3, a, N);
  initWith(4, b, N);
  initWith(0, c, N);

  size_t threadsPerBlock;
  size_t numberOfBlocks;

  /*
   * nsys should register performance changes when execution configuration
   * is updated.
   */

  threadsPerBlock = 256;
  numberOfBlocks = 6;

  cudaError_t addVectorsErr;
  cudaError_t asyncErr;

  addVectorsInto<<<numberOfBlocks, threadsPerBlock>>>(c, a, b, N);

  addVectorsErr = cudaGetLastError();
  if(addVectorsErr != cudaSuccess) printf("Error: %s\n", cudaGetErrorString(addVectorsErr));

  asyncErr = cudaDeviceSynchronize();
  if(asyncErr != cudaSuccess) printf("Error: %s\n", cudaGetErrorString(asyncErr));

  checkElementsAre(7, c, N);

  cudaFree(a);
  cudaFree(b);
  cudaFree(c);
}

CMakeLists.txt文件

cmake_minimum_required(VERSION 3.0.2)
project(learn_cuda)

## Compile as C++11, supported in ROS Kinetic and newer
# add_compile_options(-std=c++11)

## Find catkin macros and libraries
## if COMPONENTS list like find_package(catkin REQUIRED COMPONENTS xyz)
## is used, also find other catkin packages
find_package( CUDA REQUIRED )
find_package(catkin REQUIRED COMPONENTS
  roscpp
  rospy
  std_msgs
)

set(CUDA_NVCC_FLAGS "${CUDA_NVCC_FLAGS} -DMY_DEF=1")
set( CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -DMY_DEF=1" )
set( CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -DMY_DEF=1" )

## System dependencies are found with CMake's conventions
# find_package(Boost REQUIRED COMPONENTS system)


## Uncomment this if the package has a setup.py. This macro ensures
## modules and global scripts declared therein get installed
## See http://ros.org/doc/api/catkin/html/user_guide/setup_dot_py.html
# catkin_python_setup()

################################################
## Declare ROS messages, services and actions ##
################################################

## To declare and build messages, services or actions from within this
## package, follow these steps:
## * Let MSG_DEP_SET be the set of packages whose message types you use in
##   your messages/services/actions (e.g. std_msgs, actionlib_msgs, ...).
## * In the file package.xml:
##   * add a build_depend tag for "message_generation"
##   * add a build_depend and a exec_depend tag for each package in MSG_DEP_SET
##   * If MSG_DEP_SET isn't empty the following dependency has been pulled in
##     but can be declared for certainty nonetheless:
##     * add a exec_depend tag for "message_runtime"
## * In this file (CMakeLists.txt):
##   * add "message_generation" and every package in MSG_DEP_SET to
##     find_package(catkin REQUIRED COMPONENTS ...)
##   * add "message_runtime" and every package in MSG_DEP_SET to
##     catkin_package(CATKIN_DEPENDS ...)
##   * uncomment the add_*_files sections below as needed
##     and list every .msg/.srv/.action file to be processed
##   * uncomment the generate_messages entry below
##   * add every package in MSG_DEP_SET to generate_messages(DEPENDENCIES ...)

## Generate messages in the 'msg' folder
# add_message_files(
#   FILES
#   Message1.msg
#   Message2.msg
# )

## Generate services in the 'srv' folder
# add_service_files(
#   FILES
#   Service1.srv
#   Service2.srv
# )

## Generate actions in the 'action' folder
# add_action_files(
#   FILES
#   Action1.action
#   Action2.action
# )

## Generate added messages and services with any dependencies listed here
# generate_messages(
#   DEPENDENCIES
#   std_msgs
# )

################################################
## Declare ROS dynamic reconfigure parameters ##
################################################

## To declare and build dynamic reconfigure parameters within this
## package, follow these steps:
## * In the file package.xml:
##   * add a build_depend and a exec_depend tag for "dynamic_reconfigure"
## * In this file (CMakeLists.txt):
##   * add "dynamic_reconfigure" to
##     find_package(catkin REQUIRED COMPONENTS ...)
##   * uncomment the "generate_dynamic_reconfigure_options" section below
##     and list every .cfg file to be processed

## Generate dynamic reconfigure parameters in the 'cfg' folder
# generate_dynamic_reconfigure_options(
#   cfg/DynReconf1.cfg
#   cfg/DynReconf2.cfg
# )

###################################
## catkin specific configuration ##
###################################
## The catkin_package macro generates cmake config files for your package
## Declare things to be passed to dependent projects
## INCLUDE_DIRS: uncomment this if your package contains header files
## LIBRARIES: libraries you create in this project that dependent projects also need
## CATKIN_DEPENDS: catkin_packages dependent projects also need
## DEPENDS: system dependencies of this project that dependent projects also need
catkin_package(
#  INCLUDE_DIRS include
#  LIBRARIES test_pkg
 CATKIN_DEPENDS roscpp rospy std_msgs
#  DEPENDS system_lib
)

###########
## Build ##
###########

## Specify additional locations of header files
## Your package locations should be listed before other locations
include_directories(
# include
  ${catkin_INCLUDE_DIRS}
)

## Declare a C++ library
# add_library(${PROJECT_NAME}
#   src/${PROJECT_NAME}/test_pkg.cpp
# )

## Add cmake target dependencies of the library
## as an example, code may need to be generated before libraries
## either from message generation or dynamic reconfigure
# add_dependencies(${PROJECT_NAME} ${${PROJECT_NAME}_EXPORTED_TARGETS} ${catkin_EXPORTED_TARGETS})

## Declare a C++ executable
## With catkin_make all packages are built within a single CMake context
## The recommended prefix ensures that target names across packages don't collide
# add_executable(${PROJECT_NAME}_node src/test_pkg_node.cpp)
cuda_add_executable(cuda_add_node src/01-vector-add.cu)


## Rename C++ executable without prefix
## The above recommended prefix causes long target names, the following renames the
## target back to the shorter version for ease of user use
## e.g. "rosrun someones_pkg node" instead of "rosrun someones_pkg someones_pkg_node"
# set_target_properties(${PROJECT_NAME}_node PROPERTIES OUTPUT_NAME node PREFIX "")

## Add cmake target dependencies of the executable
## same as for the library above
# add_dependencies(${PROJECT_NAME}_node ${${PROJECT_NAME}_EXPORTED_TARGETS} ${catkin_EXPORTED_TARGETS})

## Specify libraries to link a library or executable target against
target_link_libraries(cuda_add_node
  ${catkin_LIBRARIES}
)

#############
## Install ##
#############

# all install targets should use catkin DESTINATION variables
# See http://ros.org/doc/api/catkin/html/adv_user_guide/variables.html

## Mark executable scripts (Python etc.) for installation
## in contrast to setup.py, you can choose the destination
# catkin_install_python(PROGRAMS
#   scripts/my_python_script
#   DESTINATION ${CATKIN_PACKAGE_BIN_DESTINATION}
# )

## Mark executables for installation
## See http://docs.ros.org/melodic/api/catkin/html/howto/format1/building_executables.html
# install(TARGETS ${PROJECT_NAME}_node
#   RUNTIME DESTINATION ${CATKIN_PACKAGE_BIN_DESTINATION}
# )

## Mark libraries for installation
## See http://docs.ros.org/melodic/api/catkin/html/howto/format1/building_libraries.html
# install(TARGETS ${PROJECT_NAME}
#   ARCHIVE DESTINATION ${CATKIN_PACKAGE_LIB_DESTINATION}
#   LIBRARY DESTINATION ${CATKIN_PACKAGE_LIB_DESTINATION}
#   RUNTIME DESTINATION ${CATKIN_GLOBAL_BIN_DESTINATION}
# )

## Mark cpp header files for installation
# install(DIRECTORY include/${PROJECT_NAME}/
#   DESTINATION ${CATKIN_PACKAGE_INCLUDE_DESTINATION}
#   FILES_MATCHING PATTERN "*.h"
#   PATTERN ".svn" EXCLUDE
# )

## Mark other files for installation (e.g. launch and bag files, etc.)
# install(FILES
#   # myfile1
#   # myfile2
#   DESTINATION ${CATKIN_PACKAGE_SHARE_DESTINATION}
# )

#############
## Testing ##
#############

## Add gtest based cpp test target and link libraries
# catkin_add_gtest(${PROJECT_NAME}-test test/test_test_pkg.cpp)
# if(TARGET ${PROJECT_NAME}-test)
#   target_link_libraries(${PROJECT_NAME}-test ${PROJECT_NAME})
# endif()

## Add folders to be run by python nosetests
# catkin_add_nosetests(test)

节点运行

source setup.bash之后 rosrun learn_cuda cuda_add_node 可运行成功

主要参考

https://answers.ros.org/question/162646/how-to-catkin_make-cuda-code/

find_package( CUDA REQUIRED )

 # set CUDA_NVCC_FLAGS as you would do with CXX/C FLAGS                        
  set(CUDA_NVCC_FLAGS "${CUDA_NVCC_FLAGS} -DMY_DEF=1")
  set( CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -DMY_DEF=1" )
  set( CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -DMY_DEF=1" )

 # only necessary if you wish to create a lib
  set(CUDA_ATTACH_VS_BUILD_RULE_TO_CUDA_FILE OFF)
  set(BUILD_SHARED_LIBS ON)

# cuda compile GPU_CODE for executable
  cuda_compile(  HELLO_CU_O src/hello.cu )

## create an executable
## can add other (none CUDA) cpp files here
  cuda_add_executable( hello_exec ${LINE_PROFILER_CU_O} )
  target_link_libraries( hello_exec ${catkin_LIBRARIES} )
  add_dependencies( hello_lib ${PROJECT_NAME}_generate_messages_cpp  ${LINE_PROFILER_CU_O} )

## if message headers needed 
  add_dependencies( hello_lib ${PROJECT_NAME}_generate_messages_cpp )

### or create a lib 
  cuda_add_library( ${PROJECT_NAME} src/lib_cuda_code.cu src/libe_none_cuda_code.cpp )
  target_link_libraries( ${PROJECT_NAME} ${catkin_LIBRARIES}  )

## if message headers needed 
  add_dependencies( ${PROJECT_NAME} ${PROJECT_NAME}_generate_messages_cpp )