[detect_grasps-1] process has died [pid 26762, exit code -11,

[jediofgever]

Hello, I am trying to use this lib with my simulated point cloud under gazebo,but I am getting error in title. I see the log but there is no any meaningful output for further debug. Here is what happens when Launch the node;

atas@atas-G7-7590:~/catkin_ws$ roslaunch gpd_ros ur5.launch 
... logging to /home/atas/.ros/log/9273b2be-62b2-11ea-ad7d-e4b97aeb78f3/roslaunch-atas-G7-7590-26747.log
Checking log directory for disk usage. This may take awhile.
Press Ctrl-C to interrupt
WARNING: disk usage in log directory [/home/atas/.ros/log] is over 1GB.
It's recommended that you use the 'rosclean' command.

started roslaunch server http://atas-G7-7590:36443/

SUMMARY
========

PARAMETERS
 * /detect_grasps/cloud_topic: /camera/depth/points
 * /detect_grasps/cloud_type: 0
 * /detect_grasps/config_file: /home/atas/gpd/cf...
 * /detect_grasps/rviz_topic: plot_grasps
 * /detect_grasps/samples_topic: 
 * /rosdistro: melodic
 * /rosversion: 1.14.3

NODES
  /
    detect_grasps (gpd_ros/detect_grasps)

ROS_MASTER_URI=http://localhost:11311

process[detect_grasps-1]: started with pid [26762]
Init ....
============ HAND GEOMETRY ======================
finger_width: 0.01
hand_outer_diameter: 0.12
hand_depth: 0.06
hand_height: 0.02
init_bite: 0.01
=================================================
============ PLOTTING ========================
plot_normals: false
plot_samples false
plot_candidates: false
plot_filtered_candidates: false
plot_valid_grasps: false
plot_clustered_grasps: false
plot_selected_grasps: true
==============================================
============ CLOUD PREPROCESSING =============
voxelize: true
voxel_size: 0.003
remove_outliers: false
workspace: -1.00 1.00 -1.00 1.00 -1.00 1.00 
sample_above_plane: false
normals_radius: 0.030
refine_normals_k: 0
==============================================
============ CANDIDATE GENERATION ============
num_samples: 500
num_threads: 4
nn_radius: 0.01
hand axes: 2 
num_orientations: 8
num_finger_placements: 10
deepen_hand: true
friction_coeff: 20.00
min_viable: 6
==============================================
============ GRASP IMAGE GEOMETRY ===============
volume width: 0.1
volume depth: 0.06
volume height: 0.02
image_size: 60
image_num_channels: 15
=================================================
NET SETUP runtime: 0.12472
============ CLASSIFIER ======================
model_file: 
weights_file: /home/atas/gpd/models/lenet/15channels/params/
batch_size: 1
==============================================
============ CANDIDATE FILTERING =============
candidate_workspace: -1.00 1.00 -1.00 1.00 -1.00 1.00 
min_aperture: 0.0000
max_aperture: 0.8500
==============================================
============ CLUSTERING ======================
min_inliers: 1
==============================================

Created GPD ....
[ INFO] [1583834549.041823993]: Waiting for point cloud to arrive ...
[ INFO] [1583834549.278995343, 4695.919000000]: Received cloud with 307200 points.
Processing cloud with 307200 points.
Voxelized cloud: 113000
Calculating surface normals ...
num_threads: 4
 runtime(computeNormals): 2.4907
camera: 0, #indices: 113000, #normals: 113000 
Calculated 113000 surface normals in 2.5002s (mode: OpenMP).
Reversing direction of normals that do not point to at least one camera ...
 reversed 0 normals
 runtime (reverse normals): 0.00284553
Estimating local reference frames ...
Estimated 500 frames in 0.0042s.
Finding hand poses ...
[detect_grasps-1] process has died [pid 26762, exit code -11, cmd /home/atas/catkin_ws/devel/lib/gpd_ros/detect_grasps __name:=detect_grasps __log:=/home/atas/.ros/log/9273b2be-62b2-11ea-ad7d-e4b97aeb78f3/detect_grasps-1.log].
log file: /home/atas/.ros/log/9273b2be-62b2-11ea-ad7d-e4b97aeb78f3/detect_grasps-1*.log
all processes on machine have died, roslaunch will exit
shutting down processing monitor...
... shutting down processing monitor complete
done

is there any suggestions ? Thank you for your time

[seann999]

Same here

Console:

Created GPD ....
[ INFO] [1584601310.962219117]: Waiting for point cloud to arrive ...
[ INFO] [1584601338.677122170]: Received cloud with 307200 points.
Processing cloud with 307200 points.
Cloud after removing NANs: 269020
[detect_grasps-1] process has died [pid 18329, exit code -11, cmd /root/ROB/catkin_ws/devel/lib/gpd_ros/detect_grasps __name:=detect_grasps __log:=/root/.ros/log/20200319-053710_add2b6a2-69a3-11ea-adb8-0030644a53eb/detect_grasps-1.log].
log file: /root/.ros/log/20200319-053710_add2b6a2-69a3-11ea-adb8-0030644a53eb/detect_grasps-1*.log
all processes on machine have died, roslaunch will exit
shutting down processing monitor...
... shutting down processing monitor complete
done

[seann999]

I think I solved it after following a comment here

[FrankLin9981]

@seann999 Excuse me. Could you provide some details about how you solve this problem? Thanks.

[jediofgever]

@seann999 the issue you point is related to PCL, however the output of gdb seems like error is triggered by Eigen;

Thread 1 "detect_grasps" received signal SIGSEGV, Segmentation fault.
Eigen::internal::dense_assignment_loop<Eigen::internal::generic_dense_assignment_kernel<Eigen::internal::evaluator<Eigen::Matrix<float, -1, -1, 0, -1, -1> >, Eigen::internal::evaluator<Eigen::Map<Eigen::Matrix<float, -1, -1, 1, -1, -1> const, 0, Eigen::Stride<0, 0> > >, Eigen::internal::assign_op<float, float>, 0>, 0, 0>::run (kernel=<synthetic pointer>...)
    at /usr/include/eigen3/Eigen/src/Core/AssignEvaluator.h:326
326         kernel.assignCoeffByOuterInner(outer, inner);
(gdb) BT

[jediofgever]

This issue happened to me because the network paths were not correctly found

I used Eigen classifier, which uses some parameters under lenet folder. I think @atenpas moved lenet folder from the root of project to models folder but the default path in the ros_eigen_params.cfg was not updated accordingly. Finally leading this issue. So anyone who met this issue make sure your path to the classifiers are correctly set

[darmin-hsiung]

The same question has show when I run $ roslaunch gpd tutorial1.launch This gpd code is forword version, please help me find where the question at? Thanks again. The error log are as follows

process[detect_grasps-1]: started with pid [9269] NET SETUP runtime: 0.0985822 [ INFO] [1614949591.731973322]: Waiting for point cloud to arrive ... [ INFO] [1614949591.976138311]: Received cloud with 170258 points. Processing cloud with: 170258 points. After workspace filtering: 170258 points left. double free or corruption (out) [detect_grasps-1] process has died [pid 9269, exit code -6, cmd /home/wyu/catkin_ws/devel/lib/gpd/detect_grasps __name:=detect_grasps __log:=/home/wyu/.ros/log/c8eacd48-7db2-11eb-9f63-0024d7d351e8/detect_grasps-1.log]. log file: /home/wyu/.ros/log/c8eacd48-7db2-11eb-9f63-0024d7d351e8/detect_grasps-1*.log all processes on machine have died, roslaunch will exit shutting down processing monitor... ... shutting down processing monitor complete done

[FabianSchuetze]

I was just wrapping a unique pointer around the cloud_camera object in the grasp_detection_node.cpp file. Deleting this object causes the segfault on my machine.

[zhaoguangbao]

[detect_grasps-1] process has died Make sure the path to the .cfg file is set correctly before running roslaunch. for example

# Path to config file for robot hand geometry
hand_geometry_filename = /home/zhao/git_ws/gpd/cfg/hand_geometry.cfg

# Path to config file for volume and image geometry
image_geometry_filename = /home/zhao/git_ws/gpd/cfg/image_geometry_15channels.cfg

# Path to directory that contains neural network parameters
weights_file = /home/zhao/git_ws/gpd/models/lenet/15channels/params/

Maybe "double free or corruption (out)" error will occur, which can be solved as follows

// grasp_detection_node.cpp
void GraspDetectionNode::cloud_callback(const sensor_msgs::PointCloud2& msg)
{
  if (!has_cloud_)
  {
      if(cloud_camera_)
        delete cloud_camera_;
      cloud_camera_ = NULL;
      // ...

[friolero]

[detect_grasps-1] process has died Make sure the path to the .cfg file is set correctly before running roslaunch. for example

# Path to config file for robot hand geometry
hand_geometry_filename = /home/zhao/git_ws/gpd/cfg/hand_geometry.cfg

# Path to config file for volume and image geometry
image_geometry_filename = /home/zhao/git_ws/gpd/cfg/image_geometry_15channels.cfg

# Path to directory that contains neural network parameters
weights_file = /home/zhao/git_ws/gpd/models/lenet/15channels/params/

Maybe "double free or corruption (out)" error will occur, which can be solved as follows

// grasp_detection_node.cpp
void GraspDetectionNode::cloud_callback(const sensor_msgs::PointCloud2& msg)
{
  if (!has_cloud_)
  {
      if(!cloud_camera_)
        delete cloud_camera_;
      cloud_camera_ = NULL;
      // ...

I'm experincing memory leak with adding the if (!cloudcamera) for the 'double free or corruption (out)' error. Please be careful if anyone is doing that.

Removing the '-march=native -msse4.2 -mavx2 -mfma -flto' flags in gpd CMakeLists and rebuilding gpd + gpd_ros somehow solves the issue for me. Appreciate if anyone has more insight on this issue and why this helps (or hides the problem if it is?)

[zhaoguangbao]

For memory leak, updating if(!cloud_camera_) to if(cloud_camera_) may be helpful.

[hqf2591102423]

我遇到这个问题是因为未正确找到网络路径

我使用了特征分类器，它使用lenet文件夹下的一些参数。我认为@atenpaslenet将文件夹从项目根目录 移动到models文件夹，但默认路径ros_eigen_params.cfg未相应更新。终于主导了这个问题。 因此，遇到此问题的任何人都请确保正确设置了分类器的路径

确实是这个问题