can't build segmapper - Githubissues

Hi @smauq, I encounter the same problem as #106 . I compiled and pip installed the newest TF, version 1.10.

(segmappyenv) tuan@tuan:~/segmap_ws$ pip show tensorflow
Name: tensorflow
Version: 1.10.0
Summary: TensorFlow is an open source machine learning framework for everyone.
Home-page: https://www.tensorflow.org/
Author: Google Inc.
Author-email: opensource@google.com
License: Apache 2.0
Location: /home/tuan/segmappyenv/lib/python2.7/site-packages
Requires: setuptools, astor, tensorboard, enum34, gast, keras-preprocessing, six, keras-applications, wheel, absl-py, backports.weakref, termcolor, protobuf, numpy, grpcio, mock
Required-by: 
(segmappyenv) tuan@tuan:~/segmap_ws$

But I can't build segmapper. The error stays the same

__________________________________________________________________________________________________________________
Errors     << segmapper:make /home/tuan/segmap_ws/logs/segmapper/build.make.002.log                               
/home/tuan/segmap_ws/devel/lib/libtf_graph_executor.so: undefined reference to `tensorflow::ReadBinaryProto(tensorflow::Env*, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&, google::protobuf::MessageLite*)'
/home/tuan/segmap_ws/devel/lib/libtf_graph_executor.so: undefined reference to `tensorflow::internal::CheckOpMessageBuilder::NewString[abi:cxx11]()'
collect2: error: ld returned 1 exit status
make[2]: *** [/home/tuan/segmap_ws/devel/lib/segmapper/segmapper_node] Error 1
make[1]: *** [CMakeFiles/segmapper_node.dir/all] Error 2
make: *** [all] Error 2
cd /home/tuan/segmap_ws/build/segmapper; catkin build --get-env segmapper | catkin env -si  /usr/bin/make --jobserver-fds=6,7 -j; cd -
..................................................................................................................
Failed     << segmapper:make                     [ Exited with code 2 ]                                           
Failed    <<< segmapper                          [ 3.4 seconds ]                                                  
[build] Summary: 21 of 22 packages succeeded.                                                                     
[build]   Ignored:   9 packages were skipped or are blacklisted.                                                  
[build]   Warnings:  4 packages succeeded with warnings.                                                          
[build]   Abandoned: None.                                                                                        
[build]   Failed:    1 packages failed.                                                                           
[build] Runtime: 7.2 seconds total.                                                                               
(segmappyenv) tuan@tuan:~/segmap_ws$

What should I do next? Thank you for your help!

Did you also try the parameters he used when compiling tensorflow? E.g.

bazel build --config=opt --define framework_shared_object=true tensorflow:libtensorflow_cc.so

And also if that doesn't work, maybe also try version 1.9. Here's an issue discussing what appears to be the same problem https://github.com/tensorflow/tensorflow/issues/14632.

Yes, I did use that command. However, AFAIK, since I did a pip install from .whl file, there is no need to use libtensorflow_cc.so for linker, right? Please correct me if I'm wrong. I will try version 1.9. Thank you for your help!

No, you don't need it to be linked. But the necessary .so files should be included in the python package. Another solution that should definitely work then is the following. Not linking against the pip install but building exactly with the above parameters and then setting the cmake variables to link against the bazel compiled version when doing catkin build tensorflow_ros_cpp. I can try to help with the flags if the thing with version 1.9 doesn't work. Will need to test it on my machine.

Thank you for your help. I followed the instructions in tensorflow_ros_cpp and was able to compile everything. But when I try roslaunch segmapper kitti_localization.launch, I have another issue:

(segmappyenv) tuan@tuan:~/segmap_ws$ roslaunch segmapper kitti_localization.launch 
... logging to /home/tuan/.ros/log/4aea169c-b6b3-11e8-94ee-6045cb30088c/roslaunch-tuan-9440.log
Checking log directory for disk usage. This may take awhile.
Press Ctrl-C to interrupt
Done checking log file disk usage. Usage is <1GB.

started roslaunch server http://tuan:42377/

SUMMARY
========

PARAMETERS
 * /SegMapper/Benchmarker/enable_live_output: False
 * /SegMapper/Benchmarker/results_directory: /tmp/segmap_bench...
 * /SegMapper/Benchmarker/save_statistics_only: False
 * /SegMapper/LaserSlamWorker/assembled_cloud_sub_topic: /velodyne_points
 * /SegMapper/LaserSlamWorker/create_filtered_map: True
 * /SegMapper/LaserSlamWorker/distant_map_pub_topic: distant_map
 * /SegMapper/LaserSlamWorker/full_map_pub_topic: full_map
 * /SegMapper/LaserSlamWorker/local_map_pub_topic: local_map
 * /SegMapper/LaserSlamWorker/map_publication_rate_hz: 3
 * /SegMapper/LaserSlamWorker/odom_frame: world
 * /SegMapper/LaserSlamWorker/odometry_trajectory_pub_topic: odometry_trajectory
 * /SegMapper/LaserSlamWorker/publish_distant_map: False
 * /SegMapper/LaserSlamWorker/publish_full_map: False
 * /SegMapper/LaserSlamWorker/publish_local_map: True
 * /SegMapper/LaserSlamWorker/sensor_frame: velodyne
 * /SegMapper/LaserSlamWorker/trajectory_pub_topic: trajectory
 * /SegMapper/LaserSlamWorker/use_odometry_information: True
 * /SegMapper/OnlineEstimator/LaserTrack/add_m_estimator_on_icp: False
 * /SegMapper/OnlineEstimator/LaserTrack/add_m_estimator_on_odom: False
 * /SegMapper/OnlineEstimator/LaserTrack/force_priors: False
 * /SegMapper/OnlineEstimator/LaserTrack/icp_noise_model: [0.5, 0.05, 0.05,...
 * /SegMapper/OnlineEstimator/LaserTrack/nscan_in_sub_map: 5
 * /SegMapper/OnlineEstimator/LaserTrack/nscan_to_match: 3
 * /SegMapper/OnlineEstimator/LaserTrack/odometry_noise_model: [0.5, 0.5, 0.5, 0...
 * /SegMapper/OnlineEstimator/LaserTrack/save_icp_results: False
 * /SegMapper/OnlineEstimator/LaserTrack/use_icp_factors: False
 * /SegMapper/OnlineEstimator/LaserTrack/use_odom_factors: True
 * /SegMapper/OnlineEstimator/add_m_estimator_on_loop_closures: True
 * /SegMapper/OnlineEstimator/do_icp_step_on_loop_closures: False
 * /SegMapper/OnlineEstimator/loop_closure_noise_model: [0.0005, 0.0005, ...
 * /SegMapper/OnlineEstimator/loop_closures_sub_maps_radius: 3
 * /SegMapper/SegMatchWorker/SegMatch/Classifier/apply_hard_threshold_on_feature_distance: True
 * /SegMapper/SegMatchWorker/SegMatch/Classifier/enable_two_stage_retrieval: True
 * /SegMapper/SegMatchWorker/SegMatch/Classifier/feature_distance_threshold: 0.01
 * /SegMapper/SegMatchWorker/SegMatch/Classifier/knn_feature_dim: 7
 * /SegMapper/SegMatchWorker/SegMatch/Classifier/max_eigen_features_values: [2493.5, 186681.0...
 * /SegMapper/SegMatchWorker/SegMatch/Classifier/n_nearest_neighbours: 120
 * /SegMapper/SegMatchWorker/SegMatch/Classifier/normalize_eigen_for_hard_threshold: False
 * /SegMapper/SegMatchWorker/SegMatch/Classifier/normalize_eigen_for_knn: False
 * /SegMapper/SegMatchWorker/SegMatch/Classifier/threshold_to_accept_match: 0.6
 * /SegMapper/SegMatchWorker/SegMatch/Descriptors/descriptor_types: ['EigenvalueBased']
 * /SegMapper/SegMatchWorker/SegMatch/GeometricConsistency/max_consistency_distance_for_caching: 3.0
 * /SegMapper/SegMatchWorker/SegMatch/GeometricConsistency/min_cluster_size: 6
 * /SegMapper/SegMatchWorker/SegMatch/GeometricConsistency/recognizer_type: Incremental
 * /SegMapper/SegMatchWorker/SegMatch/GeometricConsistency/resolution: 0.4
 * /SegMapper/SegMatchWorker/SegMatch/LocalMap/max_vertical_distance_m: 999.0
 * /SegMapper/SegMatchWorker/SegMatch/LocalMap/min_points_per_voxel: 1
 * /SegMapper/SegMatchWorker/SegMatch/LocalMap/min_vertical_distance_m: -999.0
 * /SegMapper/SegMatchWorker/SegMatch/LocalMap/neighbors_provider_type: KdTree
 * /SegMapper/SegMatchWorker/SegMatch/LocalMap/radius_m: 50
 * /SegMapper/SegMatchWorker/SegMatch/LocalMap/voxel_size_m: 0.1
 * /SegMapper/SegMatchWorker/SegMatch/Segmenters/max_cluster_size: 15000
 * /SegMapper/SegMatchWorker/SegMatch/Segmenters/min_cluster_size: 100
 * /SegMapper/SegMatchWorker/SegMatch/Segmenters/radius_for_growing: 0.2
 * /SegMapper/SegMatchWorker/SegMatch/Segmenters/segmenter_type: IncrementalEuclid...
 * /SegMapper/SegMatchWorker/SegMatch/boundary_radius_m: 999.0
 * /SegMapper/SegMatchWorker/SegMatch/centroid_distance_threshold_m: 2
 * /SegMapper/SegMatchWorker/SegMatch/check_pose_lies_below_segments: False
 * /SegMapper/SegMatchWorker/SegMatch/filter_boundary_segments: True
 * /SegMapper/SegMatchWorker/SegMatch/filter_duplicate_segments: True
 * /SegMapper/SegMatchWorker/SegMatch/min_time_between_segment_for_matches_s: 0
 * /SegMapper/SegMatchWorker/SegMatch/segmentation_height_above_m: 999.0
 * /SegMapper/SegMatchWorker/SegMatch/segmentation_height_below_m: 999.0
 * /SegMapper/SegMatchWorker/SegMatch/segmentation_radius_m: 50.0
 * /SegMapper/SegMatchWorker/align_target_map_on_first_loop_closure: False
 * /SegMapper/SegMatchWorker/close_loops: False
 * /SegMapper/SegMatchWorker/distance_between_segmentations_m: 0.0
 * /SegMapper/SegMatchWorker/distance_to_lower_target_cloud_for_viz_m: 25.0
 * /SegMapper/SegMatchWorker/export_segments_and_matches: False
 * /SegMapper/SegMatchWorker/line_scale_loop_closures: 5.0
 * /SegMapper/SegMatchWorker/line_scale_matches: 0.4
 * /SegMapper/SegMatchWorker/localize: True
 * /SegMapper/SegMatchWorker/publish_predicted_segment_matches: False
 * /SegMapper/SegMatchWorker/ratio_of_points_to_keep_when_publishing: 1.0
 * /SegMapper/SegMatchWorker/target_cloud_filename: /home/tuan/bags/d...
 * /SegMapper/clear_local_map_after_loop_closure: False
 * /SegMapper/number_of_robots: 1
 * /SegMapper/publish_world_to_odom: True
 * /SegMapper/robot_prefix: na
 * /SegMapper/tf_publication_rate_hz: 10
 * /SegMapper/world_frame: map
 * /rosdistro: kinetic
 * /rosversion: 1.12.13
 * /segmapper/icp_configuration_file: /home/tuan/segmap...
 * /segmapper/icp_input_filters_file: /home/tuan/segmap...
 * /use_sim_time: True

NODES
  /
    player (rosbag/play)
    segmapper (segmapper/segmapper_node)
    visualizer (rviz/rviz)

auto-starting new master
process[master]: started with pid [9452]
ROS_MASTER_URI=http://localhost:11311

setting /run_id to 4aea169c-b6b3-11e8-94ee-6045cb30088c
process[rosout-1]: started with pid [9465]
started core service [/rosout]
process[visualizer-2]: started with pid [9489]
process[player-3]: started with pid [9490]
process[segmapper-4]: started with pid [9491]
[ INFO] [1536774181.240618118]: Opening /home/tuan/bags/2011_10_03_drive_27.bag

Waiting 0.2 seconds after advertising topics...[ INFO] [1536774181.328468764]: rviz version 1.12.16
[ INFO] [1536774181.328523692]: compiled against Qt version 5.5.1
[ INFO] [1536774181.328535561]: compiled against OGRE version 1.9.0 (Ghadamon)
 done.

Hit space to toggle paused, or 's' to step.
[segmapper-4] process has died [pid 9491, exit code -11, cmd /home/tuan/segmap_ws/devel/lib/segmapper/segmapper_node __name:=segmapper __log:=/home/tuan/.ros/log/4aea169c-b6b3-11e8-94ee-6045cb30088c/segmapper-4.log].
log file: /home/tuan/.ros/log/4aea169c-b6b3-11e8-94ee-6045cb30088c/segmapper-4*.log

I inspect the log file and found the problem was

[rosmaster.master][INFO] 2018-09-12 19:43:24,413: -PUB [/move_base_simple/goal] /visualizer http://tuan:34639/
[rosmaster.threadpool][ERROR] 2018-09-12 19:43:24,415: Traceback (most recent call last):
  File "/opt/ros/kinetic/lib/python2.7/dist-packages/rosmaster/threadpool.py", line 218, in run
    result = cmd(*args)
  File "/opt/ros/kinetic/lib/python2.7/dist-packages/rosmaster/master_api.py", line 210, in publisher_update_task
    ret = xmlrpcapi(api).publisherUpdate('/master', topic, pub_uris)
  File "/usr/lib/python2.7/xmlrpclib.py", line 1243, in __call__
    return self.__send(self.__name, args)
  File "/usr/lib/python2.7/xmlrpclib.py", line 1602, in __request
    verbose=self.__verbose
  File "/usr/lib/python2.7/xmlrpclib.py", line 1283, in request
    return self.single_request(host, handler, request_body, verbose)
  File "/usr/lib/python2.7/xmlrpclib.py", line 1316, in single_request
    return self.parse_response(response)
  File "/usr/lib/python2.7/xmlrpclib.py", line 1493, in parse_response
    return u.close()
  File "/usr/lib/python2.7/xmlrpclib.py", line 800, in close
    raise Fault(**self._stack[0])
Fault: <Fault -1: 'publisherUpdate: unknown method name'>

which points to this. I have no idea why this happens. Thank you for your patience!

No worries. Mind sharing what flags you used? It would be a good addition to the FAQ.

As for the error. That just means the worker thread died, not an issue with ros, but some error occured that doesn't output an error message. First things would be do you have the bag and localization maps downloaded?

@smauq Certainly, though I am not sure what flags you mean. Do you mean the flags when I compiled tensorflow? For the second question, I was able to install segmappy and downloaded all the files using segmappy_download_datasets.

tuan@tuan:~/.segmap/kitti$ ls
drive27_target_map.pcd  kitti_localization.rviz  kitti_loop_closure.rviz

I put the 2 bags file in a different folder and change the path accordingly in the launch files.

@tuandle The cmake flags you used for tensorflow_ros_cpp.

Mind posting the modified launch file and the paths to where you have moved the files? For example I see in the config that it's looking for the pcd map in folder /SegMapper/SegMatchWorker/target_cloud_filename: /home/tuan/bags/d..., but what you posted above shows the map is in ~/.segmap/kitti/drive27_target_map.pcd.

@smauq My cmake flags are as follows:

(segmappyenv) tuan@tuan:~/segmappyenv$ catkin build tensorflow_ros_cpp --cmake-args -DFORCE_TF_PIP_SEARCH="OFF" -DFORCE_TF_BAZEL_SEARCH="ON" -DTF_BAZEL_LIBRARY="/home/tuan/tensorflow/bazel-bin/tensorflow/libtensorflow_cc.so" -DTF_BAZEL_SRC_DIR="/home/tuan/tensorflow"

I did catkin clean -y to rebuild from scratch:

(segmappyenv) tuan@tuan:~/segmap_ws$ catkin build tensorflow_ros_cpp --cmake-args -DFORCE_TF_PIP_SEARCH="OFF" -DFORCE_TF_BAZEL_SEARCH="ON" -DTF_BAZEL_LIBRARY="/home/tuan/tensorflow/bazel-bin/tensorflow/libtensorflow_cc.so" -DTF_BAZEL_SRC_DIR="/home/tuan/tensorflow"
-------------------------------------------------------------------------------------------------------------------
Profile:                     default
Extending:             [env] /opt/ros/kinetic
Workspace:                   /home/tuan/segmap_ws
-------------------------------------------------------------------------------------------------------------------
Source Space:       [exists] /home/tuan/segmap_ws/src
Log Space:         [missing] /home/tuan/segmap_ws/logs
Build Space:        [exists] /home/tuan/segmap_ws/build
Devel Space:        [exists] /home/tuan/segmap_ws/devel
Install Space:      [unused] /home/tuan/segmap_ws/install
DESTDIR:            [unused] None
-------------------------------------------------------------------------------------------------------------------
Devel Space Layout:          merged
Install Space Layout:        None
-------------------------------------------------------------------------------------------------------------------
Additional CMake Args:       -DCMAKE_BUILD_TYPE=Release -DFORCE_TF_PIP_SEARCH=OFF -DFORCE_TF_BAZEL_SEARCH=ON -DTF_BAZEL_LIBRARY=/home/tuan/tensorflow/bazel-bin/tensorflow/libtensorflow_cc.so -DTF_BAZEL_SRC_DIR=/home/tuan/tensorflow
Additional Make Args:        -j8
Additional catkin Make Args: None
Internal Make Job Server:    True
Cache Job Environments:      False
-------------------------------------------------------------------------------------------------------------------
Whitelisted Packages:        None
Blacklisted Packages:        None
-------------------------------------------------------------------------------------------------------------------
Workspace configuration appears valid.

NOTE: Forcing CMake to run for each package.
-------------------------------------------------------------------------------------------------------------------
[build] Found '31' packages in 0.0 seconds.                                                                       
[build] Updating package table.                                                                                   
Starting  >>> tensorflow_ros_cpp                                                                                  
__________________________________________________________________________________________________________________
Warnings   << tensorflow_ros_cpp:cmake /home/tuan/segmap_ws/logs/tensorflow_ros_cpp/build.cmake.000.log           
-- tensorflow_ros_cpp is probing the system
-- - The system uses C++11 ABI, the use of tensorflow from pip is possible, but problematic; consider compiling tensorflow yourself
-- - Not searching for Tensorflow installed by pip since the system uses C++11 ABI. Set FORCE_TF_PIP_SEARCH to force the search.
-- - Trying to find Tensorflow compiled by bazel
-- -- Found bazel-compiled libtensorflow_cc.so, using it.
-- -- Created tensorflow library link /home/tuan/tensorflow/bazel-bin/tensorflow/libtensorflow_cc.so -> /home/tuan/segmap_ws/devel/lib/libtensorflow_cc.so.
-- -- Found Tensorflow sources dir /home/tuan/tensorflow.
-- -- The Tensorflow library is compiled with CUDA support.
-- -- Using protobuf compiler libprotoc 3.6.0
, you should compile your code with the same version of protoc.
-- -- You can do it by using 'export PATH=/home/tuan/tensorflow/bazel-out/host/bin/external/protobuf_archive:$PATH'
-- - Not searching for Tensorflow from tensorflow_catkin since Tensorflow has already been found. Set FORCE_TF_CATKIN_SEARCH to force the search.
-- Using Tensorflow library found by bazel
-- The found Tensorflow library uses C++11 ABI.
-- The found Tensorflow library is compiled with CUDA support.
cd /home/tuan/segmap_ws/build/tensorflow_ros_cpp; catkin build --get-env tensorflow_ros_cpp | catkin env -si  /usr/bin/cmake /home/tuan/segmap_ws/src/tensorflow_ros_cpp --no-warn-unused-cli -DCATKIN_DEVEL_PREFIX=/home/tuan/segmap_ws/devel -DCMAKE_INSTALL_PREFIX=/home/tuan/segmap_ws/install -DCMAKE_BUILD_TYPE=Release -DFORCE_TF_PIP_SEARCH=OFF -DFORCE_TF_BAZEL_SEARCH=ON -DTF_BAZEL_LIBRARY=/home/tuan/tensorflow/bazel-bin/tensorflow/libtensorflow_cc.so -DTF_BAZEL_SRC_DIR=/home/tuan/tensorflow; cd -
..................................................................................................................
Finished  <<< tensorflow_ros_cpp                [ 1.4 seconds ]                                                   
[build] Summary: All 1 packages succeeded!                                                                        
[build]   Ignored:   30 packages were skipped or are blacklisted.                                                 
[build]   Warnings:  1 packages succeeded with warnings.                                                          
[build]   Abandoned: None.                                                                                        
[build]   Failed:    None.                                                                                        
[build] Runtime: 1.4 seconds total.                                                                               
[build] Note: Workspace packages have changed, please re-source setup files to use them.

The kitti_localization.launch is as follows:

<?xml version="1.0" encoding="utf-8"?>

<launch>

  <arg name="respawn_value" default="false" /> <!-- boolean: true, false -->

  <!-- Bag file path -->  
  <arg name="bag_file" value="/home/tuan/bags/2011_10_03_drive_27.bag"/>

  <!-- Rviz visualization config file path -->
  <arg name="rviz_config_file" value="$(env HOME)/.segmap/kitti/kitti_localization.rviz"/>

  <!-- ROS parameters -->
  <rosparam>use_sim_time: true</rosparam>

  <rosparam command="load" file="$(find segmapper)/launch/kitti/kitti_localization.yaml"/>

  <!-- Target cloud -->
  <arg name="target_cloud" value="/home/tuan/bags/drive27_target_map.pcd"/>
  <rosparam param="/SegMapper/SegMatchWorker/target_cloud_filename" subst_value="True">$(arg target_cloud)</rosparam>

  <!-- Callgrind profiling option -->
  <arg name="enable_callgrind" default="true" />
  <arg name="play_rate" value="0.005" if="$(arg enable_callgrind)"/>
  <arg name="play_rate" value="1" unless="$(arg enable_callgrind)"/>

  <!-- rviz node -->
  <node name="visualizer" pkg="rviz" type="rviz" output="screen" respawn="$(arg respawn_value)" args="-d $(arg rviz_config_file)" />

  <!-- rosbag player node -->
  <node name="player" pkg="rosbag" type="play" output="screen" args=" -r $(arg play_rate) -s 333 --clock --pause $(arg bag_file)"/>

  <!-- segmapper node -->
  <include file="$(find segmapper)/launch/segmapper.launch">
    <arg name="icp_configuration_file" value="$(find segmapper)/launch/kitti/icp_dynamic_outdoor.yaml"/>
    <arg name="icp_input_filters_file" value="$(find segmapper)/launch/kitti/input_filters_outdoor.yaml"/>
    <arg name="enable_callgrind" value="$(arg enable_callgrind)"/>
  </include>

</launch>

I did copy the .pcd file to my bags folder.

tuan@tuan:~/bags$ ls | grep .pcd
drive27_target_map.pcd

Ok, seems to be a segmentation fault. I would suggest you switch to this branch https://github.com/ethz-asl/segmap/tree/post-fixes and compile in debug mode, e.g.

catkin config --cmake-args -DCMAKE_BUILD_TYPE=Debug

Afterwards, in the kitti_localization file remove these lines 1 and 2. You can start rviz and play the bag file separately.

Then go and change on this line value="" to value="gdb -ex run --args". Now when you run the launch file it should stop at the point where it segmentation faults. You can type in gdb "l" to see surrounding code lines or "bt" to see the backtrace. If you post those it would be helpful to figure out what is causing the issue.

p.s. For the localization mode you need to wait a bit while the localization map is being built and when it appears in rviz, then you can actually start playing the bag.

@smauq Thank you for your instructions. I followed them and these are results from gdb:

ROS_MASTER_URI=http://localhost:11311

process[segmapper-1]: started with pid [16513]
GNU gdb (GDB) 7.7
Copyright (C) 2014 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.  Type "show copying"
and "show warranty" for details.
This GDB was configured as "x86_64-unknown-linux-gnu".
Type "show configuration" for configuration details.
For bug reporting instructions, please see:
<http://www.gnu.org/software/gdb/bugs/>.
Find the GDB manual and other documentation resources online at:
<http://www.gnu.org/software/gdb/documentation/>.
For help, type "help".
Type "apropos word" to search for commands related to "word"...
Reading symbols from /home/tuan/segmap_ws/devel/lib/segmapper/segmapper_node...done.
Starting program: /home/tuan/segmap_ws/devel/lib/segmapper/segmapper_node __name:=segmapper __log:=/home/tuan/.ros/log/410b38cc-b807-11e8-b0b5-6045cb30088c/segmapper-1.log
[tcsetpgrp failed in terminal_inferior: Inappropriate ioctl for device]
[tcsetpgrp failed in terminal_inferior: Inappropriate ioctl for device]
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1".
[tcsetpgrp failed in terminal_inferior: Inappropriate ioctl for device]
[New Thread 0x7fff5e0a7700 (LWP 16607)]
[New Thread 0x7fff5d8a6700 (LWP 16608)]
[New Thread 0x7fff5b0a5700 (LWP 16609)]
[New Thread 0x7fff588a4700 (LWP 16610)]
[New Thread 0x7fff560a3700 (LWP 16611)]
[New Thread 0x7fff538a2700 (LWP 16612)]
[New Thread 0x7fff510a1700 (LWP 16613)]

Program received signal SIGSEGV, Segmentation fault.
0x00007ffff230de27 in gtsam::noiseModel::Constrained::Constrained(Eigen::Matrix<double, -1, 1, 0, -1, 1> const&, Eigen::Matrix<double, -1, 1, 0, -1, 1> const&) () from /home/tuan/segmap_ws/devel/lib/libgtsam.so.4

Output of (gdb) l:

(gdb) l
1   #include <thread>
2   
3   #include <ros/ros.h>
4   
5   #include "segmapper/segmapper.hpp"
6   
7   int main(int argc, char **argv) {
8     ros::init(argc, argv, "SegMapper");
9     ros::NodeHandle node_handle("~");
10  
(gdb) 
11    SegMapper mapper(node_handle);
12  
13    std::thread publish_map_thread(&SegMapper::publishMapThread, &mapper);
14    std::thread publish_tf_thread(&SegMapper::publishTfThread, &mapper);
15    std::thread segmatch_thread(&SegMapper::segMatchThread, &mapper);
16  
17    try {
18      ros::spin();
19    }
20    catch (const std::exception& e) {
(gdb) 
21      ROS_ERROR_STREAM("Exception: " << e.what());
22      return 1;
23    }
24    catch (...) {
25      ROS_ERROR_STREAM("Unknown Exception");
26      return 1;
27    }
28  
29    publish_map_thread.join();
30    publish_tf_thread.join();
(gdb) 
31    segmatch_thread.join();
32  
33    return 0;
34  }
(gdb) 
Line number 35 out of range; /home/tuan/segmap_ws/src/segmap/segmapper/src/segmapper_node.cpp has 34 lines.

Output of (gdb) backtrace:

(gdb) bt
#0  0x00007ffff230de27 in gtsam::noiseModel::Constrained::Constrained(Eigen::Matrix<double, -1, 1, 0, -1, 1> const&, Eigen::Matrix<double, -1, 1, 0, -1, 1> const&) () from /home/tuan/segmap_ws/devel/lib/libgtsam.so.4
#1  0x00007ffff230de8e in gtsam::noiseModel::Constrained::MixedSigmas(Eigen::Matrix<double, -1, 1, 0, -1, 1> const&, Eigen::Matrix<double, -1, 1, 0, -1, 1> const&) () from /home/tuan/segmap_ws/devel/lib/libgtsam.so.4
#2  0x00007ffff230e0e3 in gtsam::noiseModel::Diagonal::Sigmas(Eigen::Matrix<double, -1, 1, 0, -1, 1> const&, bool)
    () from /home/tuan/segmap_ws/devel/lib/libgtsam.so.4
#3  0x00007ffff21fa225 in _GLOBAL__sub_I_lago.cpp () from /home/tuan/segmap_ws/devel/lib/libgtsam.so.4
#4  0x00007ffff7de76ba in ?? () from /lib64/ld-linux-x86-64.so.2
#5  0x00007ffff7de77cb in ?? () from /lib64/ld-linux-x86-64.so.2
#6  0x00007ffff7dd7c6a in ?? () from /lib64/ld-linux-x86-64.so.2
#7  0x0000000000000003 in ?? ()
#8  0x00007fffffffd814 in ?? ()
#9  0x00007fffffffd84c in ?? ()
#10 0x00007fffffffd85e in ?? ()
#11 0x0000000000000000 in ?? ()

I will try on a different computer and see if I can make it work.

@smauq any progress on debugging this problem? I have successfully installed and run your demos on another laptop. I used TF version 1.8 on the second laptop though. But looking as gdb output it says that I have a problem with libgtsam on my first laptop. Could I do anything to help you debug this?

Not sure, seems like gtsam basically crashes, the code of which we haven't modified. Can you give more details on your system setup on which it doesn't work (OS, CPU model, amount of RAM, gcc version, if it's any specific platform/server etc.)?

What you could try to do is go into src/gtsam_catkin. If you type git status you should see something like HEAD detached at ecf6a04. Now try checking out the master (git checkout master) and recompiling. This should bring you to the newest version of gtsam and maybe the error is something that has been fixed more recently (e.g. last year and a half). Will think of more things in the mean time.

@smauq Certainly, here is my system info of my personal laptop, on which segmap crashes:

tuan@tuan:~$ sudo lshw -short
H/W path         Device     Class          Description
======================================================
                            system         GL502VSK
/0                          bus            GL502VSK
/0/0                        memory         64KiB BIOS
/0/8                        memory         32GiB System Memory
/0/8/0                      memory         16GiB SODIMM Synchronous 2400 MHz (0,4 ns)
/0/8/1                      memory         [empty]
/0/8/2                      memory         16GiB SODIMM Synchronous 2400 MHz (0,4 ns)
/0/8/3                      memory         [empty]
/0/e                        memory         256KiB L1 cache
/0/f                        memory         1MiB L2 cache
/0/10                       memory         6MiB L3 cache
/0/11                       processor      Intel(R) Core(TM) i7-7700HQ CPU @ 2.80GHz

tuan@tuan:~$ lsb_release -a
No LSB modules are available.
Distributor ID: Ubuntu
Description:    Ubuntu 16.04.4 LTS
Release:    16.04
Codename:   xenial

tuan@tuan:~$ gcc --version
gcc (Ubuntu 5.4.0-6ubuntu1~16.04.10) 5.4.0 20160609
Copyright (C) 2015 Free Software Foundation, Inc.
This is free software; see the source for copying conditions.  There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.

After switching to master branch for gtsam_catkin, I did a catkin clean and recompiled everything but the error about libgtsam is still the same.

tuan@tuan:~/segmap_ws/src/gtsam_catkin$ git status
On branch master
Your branch is up-to-date with 'origin/master'.
nothing to commit, working directory clean

Program received signal SIGSEGV, Segmentation fault.
0x00007ffff23065b7 in gtsam::noiseModel::Constrained::Constrained(Eigen::Matrix<double, -1, 1, 0, -1, 1> const&, Eigen::Matrix<double, -1, 1, 0, -1, 1> const&) () from /home/tuan/segmap_ws/devel/lib/libgtsam.so.4
(gdb) bt
#0  0x00007ffff23065b7 in gtsam::noiseModel::Constrained::Constrained(Eigen::Matrix<double, -1, 1, 0, -1, 1> const&, Eigen::Matrix<double, -1, 1, 0, -1, 1> const&) () from /home/tuan/segmap_ws/devel/lib/libgtsam.so.4
#1  0x00007ffff230661e in gtsam::noiseModel::Constrained::MixedSigmas(Eigen::Matrix<double, -1, 1, 0, -1, 1> const&, Eigen::Matrix<double, -1, 1, 0, -1, 1> const&) () from /home/tuan/segmap_ws/devel/lib/libgtsam.so.4
#2  0x00007ffff2306873 in gtsam::noiseModel::Diagonal::Sigmas(Eigen::Matrix<double, -1, 1, 0, -1, 1> const&, bool) () from /home/tuan/segmap_ws/devel/lib/libgtsam.so.4
#3  0x00007ffff21f2d45 in _GLOBAL__sub_I_lago.cpp () from /home/tuan/segmap_ws/devel/lib/libgtsam.so.4
#4  0x00007ffff7de76ba in ?? () from /lib64/ld-linux-x86-64.so.2
#5  0x00007ffff7de77cb in ?? () from /lib64/ld-linux-x86-64.so.2
#6  0x00007ffff7dd7c6a in ?? () from /lib64/ld-linux-x86-64.so.2
#7  0x0000000000000003 in ?? ()
#8  0x00007fffffffd8a0 in ?? ()
#9  0x00007fffffffd8d8 in ?? ()
#10 0x00007fffffffd8ea in ?? ()
#11 0x0000000000000000 in ?? ()

Nothing jumps out about the system config.

Let's figure out if the the error in gtsam can be more specific. You'll need to change Release to Debug on this line https://github.com/ethz-asl/gtsam_catkin/blob/master/CMakeLists.txt#L17 and then also change gtsam and gtsam_unstable to gtsamDebug and gtsam_unstableDebug respectively on this line https://github.com/ethz-asl/gtsam_catkin/blob/master/CMakeLists.txt#L43. This should then compile the entire gtsam package in debug mode and we'll see more specifically where the error is.

@smauq I made changes as you suggested. Here is the output from gdb backtrace:

segmapper_node: /usr/include/eigen3/Eigen/src/Core/Assign_MKL.h:134: static void Eigen::internal::Assignment<DstXprType, Eigen::CwiseUnaryOp<Eigen::internal::scalar_square_op<double>, SrcXprNested>, Eigen::internal::assign_op<double, double>, Eigen::internal::Dense2Dense, typename Eigen::internal::enable_if<Eigen::internal::vml_assign_traits<Dst, Src>::EnableVml>::type>::run(DstXprType&, const SrcXprType&, const Eigen::internal::assign_op<double, double>&) [with DstXprType = Eigen::Matrix<double, -1, 1>; SrcXprNested = const Eigen::ArrayWrapper<Eigen::Matrix<double, -1, 1> >; Eigen::internal::Assignment<DstXprType, Eigen::CwiseUnaryOp<Eigen::internal::scalar_square_op<double>, SrcXprNested>, Eigen::internal::assign_op<double, double>, Eigen::internal::Dense2Dense, typename Eigen::internal::enable_if<Eigen::internal::vml_assign_traits<Dst, Src>::EnableVml>::type>::SrcXprType = Eigen::CwiseUnaryOp<Eigen::internal::scalar_square_op<double>, const Eigen::ArrayWrapper<Eigen::Matrix<double, -1, 1> > >]: Assertion `dst.rows() == src.rows() && dst.cols() == src.cols()' failed.

Program received signal SIGABRT, Aborted.
0x00007fffae29f428 in raise () from /lib/x86_64-linux-gnu/libc.so.6
(gdb) bt
#0  0x00007fffae29f428 in raise () from /lib/x86_64-linux-gnu/libc.so.6
#1  0x00007fffae2a102a in abort () from /lib/x86_64-linux-gnu/libc.so.6
#2  0x00007fffae297bd7 in ?? () from /lib/x86_64-linux-gnu/libc.so.6
#3  0x00007fffae297c82 in __assert_fail () from /lib/x86_64-linux-gnu/libc.so.6
#4  0x00007ffff1c70f00 in Eigen::internal::Assignment<Eigen::Matrix<double, -1, 1, 0, -1, 1>, Eigen::CwiseUnaryOp<Eigen::internal::scalar_square_op<double>, Eigen::ArrayWrapper<Eigen::Matrix<double, -1, 1, 0, -1, 1> > const>, Eigen::internal::assign_op<double, double>, Eigen::internal::Dense2Dense, void>::run (dst=..., src=...)
    at /usr/include/eigen3/Eigen/src/Core/Assign_MKL.h:134
#5  0x00007ffff1c6f99e in Eigen::internal::call_assignment_no_alias<Eigen::Matrix<double, -1, 1, 0, -1, 1>, Eigen::CwiseUnaryOp<Eigen::internal::scalar_square_op<double>, Eigen::ArrayWrapper<Eigen::Matrix<double, -1, 1, 0, -1, 1> > const>, Eigen::internal::assign_op<double, double> > (dst=..., src=..., func=...)
    at /usr/include/eigen3/Eigen/src/Core/AssignEvaluator.h:819
#6  0x00007ffff1c6dfda in Eigen::PlainObjectBase<Eigen::Matrix<double, -1, 1, 0, -1, 1> >::_set_noalias<Eigen::CwiseUnaryOp<Eigen::internal::scalar_square_op<double>, Eigen::ArrayWrapper<Eigen::Matrix<double, -1, 1, 0, -1, 1> > const> > (this=0xc10290, other=...) at /usr/include/eigen3/Eigen/src/Core/PlainObjectBase.h:728
#7  0x00007ffff1c6bb61 in Eigen::PlainObjectBase<Eigen::Matrix<double, -1, 1, 0, -1, 1> >::_init1<Eigen::CwiseUnaryOp<Eigen::internal::scalar_square_op<double>, Eigen::ArrayWrapper<Eigen::Matrix<double, -1, 1, 0, -1, 1> > const>, Eigen::CwiseUnaryOp<Eigen::internal::scalar_square_op<double>, Eigen::ArrayWrapper<Eigen::Matrix<double, -1, 1, 0, -1, 1> > const> > (this=0xc10290, other=...) at /usr/include/eigen3/Eigen/src/Core/PlainObjectBase.h:812
#8  0x00007ffff1c69745 in Eigen::Matrix<double, -1, 1, 0, -1, 1>::Matrix<Eigen::CwiseUnaryOp<Eigen::internal::scalar_square_op<double>, Eigen::ArrayWrapper<Eigen::Matrix<double, -1, 1, 0, -1, 1> > const> > (this=0xc10290, x=...)
    at /usr/include/eigen3/Eigen/src/Core/Matrix.h:296
#9  0x00007ffff1c62c88 in gtsam::noiseModel::Diagonal::Diagonal (this=0xc10240, sigmas=...)
    at /home/tuan/segmap_ws/build/gtsam_catkin/gtsam_src-prefix/src/gtsam_src/gtsam/linear/NoiseModel.cpp:248
#10 0x00007ffff1c634ee in gtsam::noiseModel::Constrained::Constrained (this=0xc10240, mu=..., sigmas=...)
    at /home/tuan/segmap_ws/build/gtsam_catkin/gtsam_src-prefix/src/gtsam_src/gtsam/linear/NoiseModel.cpp:334
#11 0x00007ffff1c635a5 in gtsam::noiseModel::Constrained::MixedSigmas (mu=..., sigmas=...)
    at /home/tuan/segmap_ws/build/gtsam_catkin/gtsam_src-prefix/src/gtsam_src/gtsam/linear/NoiseModel.cpp:341
#12 0x00007ffff1c00d40 in gtsam::noiseModel::Constrained::MixedSigmas (sigmas=...)
    at /home/tuan/segmap_ws/build/gtsam_catkin/gtsam_src-prefix/src/gtsam_src/gtsam/linear/NoiseModel.h:419
#13 0x00007ffff1c62f51 in gtsam::noiseModel::Diagonal::Sigmas (sigmas=..., smart=true)
    at /home/tuan/segmap_ws/build/gtsam_catkin/gtsam_src-prefix/src/gtsam_src/gtsam/linear/NoiseModel.cpp:271
#14 0x00007ffff1dd868e in __static_initialization_and_destruction_0 (__initialize_p=1, __priority=65535)
    at /home/tuan/segmap_ws/build/gtsam_catkin/gtsam_src-prefix/src/gtsam_src/gtsam/slam/lago.cpp:37
#15 0x00007ffff1dd896c in _GLOBAL__sub_I_lago.cpp(void) ()
    at /home/tuan/segmap_ws/build/gtsam_catkin/gtsam_src-prefix/src/gtsam_src/gtsam/slam/lago.cpp:399
#16 0x00007ffff7de76ba in ?? () from /lib64/ld-linux-x86-64.so.2
#17 0x00007ffff7de77cb in ?? () from /lib64/ld-linux-x86-64.so.2
#18 0x00007ffff7dd7c6a in ?? () from /lib64/ld-linux-x86-64.so.2
#19 0x0000000000000003 in ?? ()
#20 0x00007fffffffd8c0 in ?? ()
#21 0x00007fffffffd8f8 in ?? ()
#22 0x00007fffffffd90a in ?? ()
#23 0x0000000000000000 in ?? ()

Ok I think I found the issue now. Seems very likely to have been a bug in eigen3 that was fixed by this commit https://bitbucket.org/eigen/eigen/commits/d144ceb83119af2ddc5b829727b5eee83b239a6a.

Options would be:

Disable MKL support (either by uninstalling mkl or through a flag)
Pull the master version of Eigen and do a system install with that (not highly recommended as it might cause confusion with the one installed from apt)
Download eigen through eigen_catkin. Set this flag to "OFF" https://github.com/ethz-asl/eigen_catkin/blob/master/use-system-installation-of-eigen.cmake. (Try to compile now and see if it works, very likely not, but it's worth testing.) Afterwards change the source repo to https://bitbucket.org/eigen/eigen/get/7eb9a67422ff.zip here https://github.com/ethz-asl/eigen_catkin/blob/master/CMakeLists.txt#L88 (i.e. pull the master branch of eigen). And also remove this line with the patch command https://github.com/ethz-asl/eigen_catkin/blob/master/CMakeLists.txt#L89

I would start with the last option and afterwards see how option one would be possible if that doesn't work out.

@smauq I followed option 3, e.g set OFF to USE_SYSTEM_EIGEN, change the link to eigen master branch, remove the patch. I didn't disable MKL though. I was able to get segmapper loaded target map :) yay However, another segfault appears with Eigen again

Reelimintated: 4  Relinearized: 2  Cliques: 4
*** stack smashing detected ***: /home/tuan/segmap_ws/devel/lib/segmapper/segmapper_node terminated

Program received signal SIGABRT, Aborted.
0x00007fffa9dc9428 in raise () from /lib/x86_64-linux-gnu/libc.so.6
(gdb) bt
#0  0x00007fffa9dc9428 in raise () from /lib/x86_64-linux-gnu/libc.so.6
#1  0x00007fffa9dcb02a in abort () from /lib/x86_64-linux-gnu/libc.so.6
#2  0x00007fffa9e0b7ea in ?? () from /lib/x86_64-linux-gnu/libc.so.6
#3  0x00007fffa9ead15c in __fortify_fail () from /lib/x86_64-linux-gnu/libc.so.6
#4  0x00007fffa9ead100 in __stack_chk_fail () from /lib/x86_64-linux-gnu/libc.so.6
#5  0x00007fffed91f159 in Eigen::DenseBase<Eigen::CwiseUnaryOp<Eigen::internal::scalar_abs_op<double>, Eigen::CwiseBinaryOp<Eigen::internal::scalar_difference_op<double, double>, Eigen::Matrix<double, -1, 1, 0, -1, 1> const, Eigen::Matrix<double, -1, 1, 0, -1, 1> const> const> >::redux<Eigen::internal::scalar_max_op<double, double> > (
    this=0x7fffffff8c70, func=...) at /home/tuan/segmap_ws/devel/include/eigen3/Eigen/src/Core/Redux.h:417
#6  0x00007fffed91dd87 in Eigen::DenseBase<Eigen::CwiseUnaryOp<Eigen::internal::scalar_abs_op<double>, Eigen::CwiseBinaryOp<Eigen::internal::scalar_difference_op<double, double>, Eigen::Matrix<double, -1, 1, 0, -1, 1> const, Eigen::Matrix<double, -1, 1, 0, -1, 1> const> const> >::maxCoeff (this=0x7fffffff8c70)
    at /home/tuan/segmap_ws/devel/include/eigen3/Eigen/src/Core/Redux.h:436
#7  0x00007fffed91bc27 in Eigen::internal::lpNorm_selector<Eigen::CwiseBinaryOp<Eigen::internal::scalar_difference_op<double, double>, Eigen::Matrix<double, -1, 1, 0, -1, 1> const, Eigen::Matrix<double, -1, 1, 0, -1, 1> const>, -1>::run (m=...) at /home/tuan/segmap_ws/devel/include/eigen3/Eigen/src/Core/Dot.h:244
#8  0x00007fffed919964 in Eigen::MatrixBase<Eigen::CwiseBinaryOp<Eigen::internal::scalar_difference_op<double, double>, Eigen::Matrix<double, -1, 1, 0, -1, 1> const, Eigen::Matrix<double, -1, 1, 0, -1, 1> const> >::lpNorm<-1> (
    this=0x7fffffff8f10) at /home/tuan/segmap_ws/devel/include/eigen3/Eigen/src/Core/Dot.h:269
#9  0x00007fffed917384 in gtsam::internal::optimizeWildfireNode<gtsam::ISAM2Clique> (clique=..., threshold=0.001, 
    changed=..., replaced=..., delta=..., count=@0x7fffffff9278: 4)
    at /home/tuan/segmap_ws/build/gtsam_catkin/gtsam_src-prefix/src/gtsam_src/gtsam/nonlinear/ISAM2-inl.h:225
#10 0x00007fffed914f2d in gtsam::optimizeWildfireNonRecursive<gtsam::ISAM2Clique> (root=..., threshold=0.001, 
    keys=..., delta=...)
    at /home/tuan/segmap_ws/build/gtsam_catkin/gtsam_src-prefix/src/gtsam_src/gtsam/nonlinear/ISAM2-inl.h:277
#11 0x00007fffed912e8a in gtsam::ISAM2::Impl::UpdateGaussNewtonDelta (roots=..., replacedKeys=..., delta=..., 
    wildfireThreshold=0.001)
    at /home/tuan/segmap_ws/build/gtsam_catkin/gtsam_src-prefix/src/gtsam_src/gtsam/nonlinear/ISAM2-impl.cpp:292
#12 0x00007fffed92d472 in gtsam::ISAM2::updateDelta (this=0x15a0900, forceFullSolve=false)
    at /home/tuan/segmap_ws/build/gtsam_catkin/gtsam_src-prefix/src/gtsam_src/gtsam/nonlinear/ISAM2.cpp:972
#13 0x00007fffed9299c7 in gtsam::ISAM2::update (this=0x15a0900, newFactors=..., newTheta=..., 
    removeFactorIndices=..., constrainedKeys=..., noRelinKeys=..., extraReelimKeys=..., force_relinearize=false)
    at /home/tuan/segmap_ws/build/gtsam_catkin/gtsam_src-prefix/src/gtsam_src/gtsam/nonlinear/ISAM2.cpp:544
#14 0x00007ffff0e043c5 in laser_slam::IncrementalEstimator::estimate (this=0x15a08a0, new_factors=..., 
    new_values=..., timestamp_ns=310854000)
    at /home/tuan/segmap_ws/src/laser_slam/laser_slam/src/incremental_estimator.cpp:158
#15 0x00007ffff3a7c251 in laser_slam_ros::LaserSlamWorker::scanCallback (this=0x15a3150, cloud_msg_in=...)
    at /home/tuan/segmap_ws/src/laser_slam/laser_slam_ros/src/laser_slam_worker.cpp:169
#16 0x00007ffff3ad9432 in boost::_mfi::mf1<void, laser_slam_ros::LaserSlamWorker, sensor_msgs::PointCloud2_<std::allocator<void> > const&>::operator() (this=0x15b3310, p=0x15a3150, a1=...)
    at /usr/include/boost/bind/mem_fn_template.hpp:165
#17 0x00007ffff3ad1f02 in boost::_bi::list2<boost::_bi::value<laser_slam_ros::LaserSlamWorker*>, boost::arg<1> >::operator()<boost::_mfi::mf1<void, laser_slam_ros::LaserSlamWorker, sensor_msgs::PointCloud2_<std::allocator<void> > ---Type <return> to continue, or q <return> to quit---
const&>, boost::_bi::list1<sensor_msgs::PointCloud2_<std::allocator<void> > const&> > (this=0x15b3320, f=..., 
    a=...) at /usr/include/boost/bind/bind.hpp:313
#18 0x00007ffff3ac90bd in boost::_bi::bind_t<void, boost::_mfi::mf1<void, laser_slam_ros::LaserSlamWorker, sensor_msgs::PointCloud2_<std::allocator<void> > const&>, boost::_bi::list2<boost::_bi::value<laser_slam_ros::LaserSlamWorker*>, boost::arg<1> > >::operator()<sensor_msgs::PointCloud2_<std::allocator<void> > const&> (this=0x15b3310, 
    a1=...) at /usr/include/boost/bind/bind.hpp:905
#19 0x00007ffff3abfb4d in boost::detail::function::void_function_obj_invoker1<boost::_bi::bind_t<void, boost::_mfi::mf1<void, laser_slam_ros::LaserSlamWorker, sensor_msgs::PointCloud2_<std::allocator<void> > const&>, boost::_bi::list2<boost::_bi::value<laser_slam_ros::LaserSlamWorker*>, boost::arg<1> > >, void, sensor_msgs::PointCloud2_<std::allocator<void> > const&>::invoke (function_obj_ptr=..., a0=...)
    at /usr/include/boost/function/function_template.hpp:159
#20 0x00007ffff3afa323 in boost::function1<void, sensor_msgs::PointCloud2_<std::allocator<void> > const&>::operator() (this=0x15b3308, a0=...) at /usr/include/boost/function/function_template.hpp:773
#21 0x00007ffff3af7719 in ros::SubscriptionCallbackHelperT<sensor_msgs::PointCloud2_<std::allocator<void> > const&, void>::call (this=0x15b3300, params=...) at /opt/ros/kinetic/include/ros/subscription_callback_helper.h:144
#22 0x00007fffc541ae2d in ros::SubscriptionQueue::call() () from /opt/ros/kinetic/lib/libroscpp.so
#23 0x00007fffc53c06f8 in ros::CallbackQueue::callOneCB(ros::CallbackQueue::TLS*) ()
   from /opt/ros/kinetic/lib/libroscpp.so
#24 0x00007fffc53c20fb in ros::CallbackQueue::callAvailable(ros::WallDuration) ()
   from /opt/ros/kinetic/lib/libroscpp.so
#25 0x00007fffc541eef9 in ros::SingleThreadedSpinner::spin(ros::CallbackQueue*) ()
   from /opt/ros/kinetic/lib/libroscpp.so
#26 0x00007fffc5403edb in ros::spin() () from /opt/ros/kinetic/lib/libroscpp.so
#27 0x00000000005a8728 in main (argc=1, argv=0x7fffffffce08)
    at /home/tuan/segmap_ws/src/segmap/segmapper/src/segmapper_node.cpp:18

Since this error has nothing to do with MKL so I guess I don't have to disable it. I will give it a try anyway.

If you try disabling MKL revert back to the original versions of Eigen and maybe also gtsam. To avoid having errors from new untested features.

I am so sorry for giving a comment here,please forgive me.My ubuntu version is 16.04 too.How can I compile everything before catkin build segmapper?In this way(Custom compilation of Tensorflow using Bazel)?

@14212094 no worries. The command is:

(segmappyenv) tuan@tuan:~/segmap_ws$ catkin build tensorflow_ros_cpp --cmake-args -DFORCE_TF_PIP_SEARCH="OFF" -DFORCE_TF_BAZEL_SEARCH="ON" -DTF_BAZEL_LIBRARY="/home/tuan/tensorflow/bazel-bin/tensorflow/libtensorflow_cc.so" -DTF_BAZEL_SRC_DIR="/home/tuan/tensorflow"

You need to substitute your path. You can refer to this for more flags.

Thank you！！Did you use this command: bazel build --config=opt --define framework_shared_object=true tensorflow:libtensorflow_cc.so before pip installing from .whl?

@14212094 yes, I used that command to build the C++ library.

@tuandle Thanks! there are something that i am not clear.According to tensorflow_ros_cpp,it says that "it is no need to continue with the guide building or installing the pip package".Did you skip these two steps?

@14212094 yes, I did skip them.

@tuandle OK! After the bazel build,I diretly catkin build the tensorflow_ros_cpp --cmake-args.....It built really quickly.It skipped 30 targets and built 1 in A few seconds(I think it is unusual).Is this normal?When i try to build segmapper,it still can't build it.I think i may lose some steps which leads to the wrong build of tensorflow_ros_cpp.Please correct me and tell me some details about the steps before catkin build segmapper if it is convenient!Thank you so much~

@14212094 what is your output of

catkin build tensorflow_ros_cpp --cmake-args -DFORCE_TF_PIP_SEARCH="OFF" -DFORCE_TF_BAZEL_SEARCH="ON" -DTF_BAZEL_LIBRARY="/home/tuan/tensorflow/bazel-bin/tensorflow/libtensorflow_cc.so" -DTF_BAZEL_SRC_DIR="/home/tuan/tensorflow"

Are you sure that you substitute the correct paths?

I am a newer to segmap, my problem may naive.My understanding is that you used this command: (bazel build --config=opt --define framework_shared_object=true tensorflow:libtensorflow_cc.so) to build C++ library,and then install tensorflow-gpu 1.8 under (segmappyenv) .Lastly,using this command:(catkin build tensorflow_ros_cpp --cmake-args -DFORCE_TF_PIP_SEARCH="OFF" -DFORCE_TF_BAZEL_SEARCH="ON" -DTF_BAZEL_LIBRARY="/home/tuan/tensorflow/bazel-bin/tensorflow/libtensorflow_cc.so" -DTF_BAZEL_SRC_DIR="/home/tuan/tensorflow") to compile everything and successfully built tensorflow_ros_cpp and segmapper.Please correct me if I am wrong~ Thank you for your help!

@14212094 No worries and yes, that's all I did.

@tuandle Thanks！this is my output of catkin build tensorflow_ros_cpp: （Finished <<< tensorflow_ros_cpp [ 3.3 seconds ]
[build] Summary: All 1 packages succeeded!
[build] Ignored: 30 packages were skipped or are blacklisted.
[build] Warnings: 1 packages succeeded with warnings.
[build] Abandoned: None.
[build] Failed: None.
[build] Runtime: 3.6 seconds total.
[build] Note: Workspace packages have changed, please re-source setup files to use them.） However，flowing the steps above,the same error occurs～

@14212094 can you copy and paste the error when building segmap ?

@tuandle Ok! Errors << tf_graph_executor:make /home/liwenyi/segmap_ws/logs/tf_graph_executor/build.make.000.log
In file included from /home/liwenyi/tensorflow/tensorflow/core/platform/tensor_coding.h:22:0, from /home/liwenyi/tensorflow/tensorflow/core/framework/resource_handle.h:19, from /home/liwenyi/tensorflow/tensorflow/core/framework/allocator.h:24, from /home/liwenyi/tensorflow/tensorflow/core/framework/tensor.h:20, from /home/liwenyi/tensorflow/tensorflow/core/public/session.h:24, from /home/liwenyi/segmap_ws/src/segmap/tf_graph_executor/src/tf_graph_executor.cpp:8: /home/liwenyi/tensorflow/tensorflow/core/lib/core/stringpiece.h:29:38: fatal error: absl/strings /string_view.h: 没有那个文件或目录 compilation terminated. make[2]: [CMakeFiles/tf_graph_executor.dir/src/tf_graph_executor.cpp.o] Error 1 make[1]: [CMakeFiles/tf_graph_executor.dir/all] Error 2 make: *** [all] Error 2

The tensorflow pagkage was cloned from github.com.How can i find this (absl/strings/string_view.h)？

@14212094 it seems the problem is with the TF version you are using. The problem was mentioned here #22240. I think you should checkout the branch r1.8 and rebuild everything.

I am sorry for writing here. I followed segmap/readme.md and this issue108 to compile segmapper. In short I have the following setup:

Ubuntu 16.04 tf-gpu-1.8.0 Python 3.5.2 Compiler GCC 5.4 Bazel 0.10.0 cuDNN 7 CUDA 9 GPU Drivers 390

I used this command for tensorflow: bazel build --config=opt --config=cuda --define framework_shared_object=false tensorflow:libtensorflow_cc.so

and installed using: pip install tensorflow-gpu==1.8.0 # not from wheel

I built the packages using:

catkin build tensorflow_ros_cpp --cmake-args -DFORCE_TF_PIP_SEARCH="OFF" -DFORCE_TF_BAZEL_SEARCH="ON" -DTF_BAZEL_LIBRARY="/home/mr/tensorflow/bazel-bin/tensorflow/libtensorflow_cc.so" -DTF_BAZEL_SRC_DIR="/home/mr/tensorflow" -DTF_PYTHON_VERSION="Python 3.5" -DTF_PIP_PATH="$HOME/segmappyenv/lib/python/site-packages/tensorflow" -DTF_PIP_EXECUTABLE="$HOME/segmappyenv/bin/pip" -DTF_PYTHON_LIBRARY="/usr/lib/x86_64-linux-gnu/libpython3.5m.so.1.0"

catkin build segmapper --cmake-args -DFORCE_TF_PIP_SEARCH="OFF" -DFORCE_TF_BAZEL_SEARCH="ON" -DTF_BAZEL_LIBRARY="/home/mr/tensorflow/bazel-bin/tensorflow/libtensorflow_cc.so" -DTF_BAZEL_SRC_DIR="/home/mr/tensorflow" -DTF_PYTHON_VERSION="Python 3.5" -DTF_PIP_PATH="$HOME/segmappyenv/lib/python/site-packages/tensorflow" -DTF_PIP_EXECUTABLE="$HOME/segmappyenv/bin/pip" -DTF_PYTHON_LIBRARY="/usr/lib/x86_64-linux-gnu/libpython3.5m.so.1.0"

But during the compilation got these errors: Failed << glog_catkin:make [ Exited with code 2 ] Failed <<< glog_catkin [ 4.1 seconds ]
Abandoned <<< eigen_checks [ Unrelated job failed ] Abandoned <<< libpointmatcher_ros [ Unrelated job failed ] Abandoned <<< minkindr [ Unrelated job failed ] Abandoned <<< minkindr_gtsam [ Unrelated job failed ] Abandoned <<< mincurves [ Unrelated job failed ] Abandoned <<< laser_slam [ Unrelated job failed ] Abandoned <<< laser_slam_ros [ Unrelated job failed ] Abandoned <<< tf_graph_executor [ Unrelated job failed ] Abandoned <<< segmatch [ Unrelated job failed ] Abandoned <<< segmatch_ros [ Unrelated job failed ] Abandoned <<< segmapper [ Unrelated job failed ]

Can you help ?

@smauq Hello, I encounter the same question with @tuandle . I built the package in a virtualenv, I used Ubuntu16.04+cuda9.0+cudnn7.0+bazel 0.10.0 and built tensorflow-r1.8 from source. When I compiled the segmap_ws package in the same virtualenv , I encounter error when I use "$ catkin build segmapper"

Errors << segmapper:make /home/james/segmap_ws/logs/segmapper/build.make.006.log /home/james/segmap_ws/devel/lib/libtf_graph_executor.so: undefined reference to `tensorflow::ReadBinaryProto(tensorflow::Env, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&, google::protobuf::MessageLite)' /home/james/segmap_ws/devel/lib/libtf_graph_executor.so: undefined reference to`tensorflow::internal::CheckOpMessageBuilder::NewString[abi:cxx11]()' collect2: error: ld returned 1 exit status make[2]: [/home/james/segmap_ws/devel/lib/segmapper/segmapper_node] Error 1 make[1]: [CMakeFiles/segmapper_node.dir/all] Error 2 make: *** [all] Error 2 cd /home/james/segmap_ws/build/segmapper; catkin build --get-env segmapper | catkin env -si /usr/bin/make --jobserver-fds=6,7 -j; cd - ............................................................................... Failed << segmapper:make [ Exited with code 2 ]
Failed <<< segmapper [ 3.5 seconds ]
[build] Summary: 21 of 22 packages succeeded.
[build] Ignored: 9 packages were skipped or are blacklisted.
[build] Warnings: 4 packages succeeded with warnings.
[build] Abandoned: None.
[build] Failed: 1 packages failed.
[build] Runtime: 13.0 seconds total.

I've seen your solution above,but I do not clearly understand .Could you please tell me how to deal with the problem? Thank you very much

Hi @smauq @tuandle thanks for your sharing! I successfully compiled segmpper under the instruction of your comments in Ubuntu16.04 with TF1,8 and CUDA9.0. However, I also encountered the segment fault with eigen lib just as the same as @tuandle . I debug the code by change the configuration of work space as Debug, adding gdb -ex run --args in the launch file and following option 3,( e.g set OFF to USE_SYSTEM_EIGEN, change the link to eigen master branch, remove the patch). The fault was thrown off as:

 (segmappyenv) lee@lee-XPS-15-9550:~/segmap_ws$ roslaunch segmapper kitti_loop_closure_debug.launch 
... logging to /home/lee/.ros/log/a4868456-be69-11e9-9c48-441ca8e3ae23/roslaunch-lee-XPS-15-9550-5640.log
Checking log directory for disk usage. This may take awhile.
Press Ctrl-C to interrupt
Done checking log file disk usage. Usage is <1GB.

started roslaunch server http://lee-XPS-15-9550:41699/

SUMMARY
========

PARAMETERS
 * /LaserMapper/bag_file: /home/lee/.segmap...
 * /SegMapper/Benchmarker/enable_live_output: False
 * /SegMapper/Benchmarker/results_directory: /tmp/segmap_bench...
 * /SegMapper/Benchmarker/save_statistics_only: False
 * /SegMapper/LaserSlamWorker/assembled_cloud_sub_topic: /velodyne_points
 * /SegMapper/LaserSlamWorker/create_filtered_map: True
 * /SegMapper/LaserSlamWorker/distant_map_pub_topic: distant_map
 * /SegMapper/LaserSlamWorker/full_map_pub_topic: full_map
 * /SegMapper/LaserSlamWorker/local_map_pub_topic: local_map
 * /SegMapper/LaserSlamWorker/map_publication_rate_hz: 3
 * /SegMapper/LaserSlamWorker/odom_frame: world
 * /SegMapper/LaserSlamWorker/odometry_trajectory_pub_topic: odometry_trajectory
 * /SegMapper/LaserSlamWorker/publish_distant_map: False
 * /SegMapper/LaserSlamWorker/publish_full_map: False
 * /SegMapper/LaserSlamWorker/publish_local_map: True
 * /SegMapper/LaserSlamWorker/sensor_frame: velodyne
 * /SegMapper/LaserSlamWorker/trajectory_pub_topic: trajectory
 * /SegMapper/LaserSlamWorker/use_odometry_information: True
 * /SegMapper/OnlineEstimator/LaserTrack/add_m_estimator_on_icp: False
 * /SegMapper/OnlineEstimator/LaserTrack/add_m_estimator_on_odom: False
 * /SegMapper/OnlineEstimator/LaserTrack/force_priors: False
 * /SegMapper/OnlineEstimator/LaserTrack/icp_noise_model: [0.5, 0.05, 0.05,...
 * /SegMapper/OnlineEstimator/LaserTrack/nscan_in_sub_map: 5
 * /SegMapper/OnlineEstimator/LaserTrack/nscan_to_match: 3
 * /SegMapper/OnlineEstimator/LaserTrack/odometry_noise_model: [0.5, 0.5, 0.5, 0...
 * /SegMapper/OnlineEstimator/LaserTrack/save_icp_results: False
 * /SegMapper/OnlineEstimator/LaserTrack/use_icp_factors: True
 * /SegMapper/OnlineEstimator/LaserTrack/use_odom_factors: True
 * /SegMapper/OnlineEstimator/add_m_estimator_on_loop_closures: True
 * /SegMapper/OnlineEstimator/do_icp_step_on_loop_closures: False
 * /SegMapper/OnlineEstimator/loop_closure_noise_model: [0.0005, 0.0005, ...
 * /SegMapper/OnlineEstimator/loop_closures_sub_maps_radius: 3
 * /SegMapper/SegMatchWorker/SegMatch/Classifier/apply_hard_threshold_on_feature_distance: True
 * /SegMapper/SegMatchWorker/SegMatch/Classifier/enable_two_stage_retrieval: True
 * /SegMapper/SegMatchWorker/SegMatch/Classifier/feature_distance_threshold: 0.01
 * /SegMapper/SegMatchWorker/SegMatch/Classifier/knn_feature_dim: 7
 * /SegMapper/SegMatchWorker/SegMatch/Classifier/max_eigen_features_values: [2493.5, 186681.0...
 * /SegMapper/SegMatchWorker/SegMatch/Classifier/n_nearest_neighbours: 120
 * /SegMapper/SegMatchWorker/SegMatch/Classifier/normalize_eigen_for_hard_threshold: True
 * /SegMapper/SegMatchWorker/SegMatch/Classifier/normalize_eigen_for_knn: False
 * /SegMapper/SegMatchWorker/SegMatch/Classifier/threshold_to_accept_match: 0.6
 * /SegMapper/SegMatchWorker/SegMatch/Descriptors/descriptor_types: ['EigenvalueBased']
 * /SegMapper/SegMatchWorker/SegMatch/GeometricConsistency/max_consistency_distance_for_caching: 3.0
 * /SegMapper/SegMatchWorker/SegMatch/GeometricConsistency/min_cluster_size: 6
 * /SegMapper/SegMatchWorker/SegMatch/GeometricConsistency/recognizer_type: Incremental
 * /SegMapper/SegMatchWorker/SegMatch/GeometricConsistency/resolution: 0.4
 * /SegMapper/SegMatchWorker/SegMatch/LocalMap/max_vertical_distance_m: 999.0
 * /SegMapper/SegMatchWorker/SegMatch/LocalMap/min_points_per_voxel: 1
 * /SegMapper/SegMatchWorker/SegMatch/LocalMap/min_vertical_distance_m: -999.0
 * /SegMapper/SegMatchWorker/SegMatch/LocalMap/neighbors_provider_type: KdTree
 * /SegMapper/SegMatchWorker/SegMatch/LocalMap/radius_m: 50
 * /SegMapper/SegMatchWorker/SegMatch/LocalMap/voxel_size_m: 0.1
 * /SegMapper/SegMatchWorker/SegMatch/Segmenters/max_cluster_size: 15000
 * /SegMapper/SegMatchWorker/SegMatch/Segmenters/min_cluster_size: 100
 * /SegMapper/SegMatchWorker/SegMatch/Segmenters/radius_for_growing: 0.2
 * /SegMapper/SegMatchWorker/SegMatch/Segmenters/segmenter_type: IncrementalEuclid...
 * /SegMapper/SegMatchWorker/SegMatch/boundary_radius_m: 999.0
 * /SegMapper/SegMatchWorker/SegMatch/centroid_distance_threshold_m: 2.5
 * /SegMapper/SegMatchWorker/SegMatch/check_pose_lies_below_segments: False
 * /SegMapper/SegMatchWorker/SegMatch/filter_boundary_segments: True
 * /SegMapper/SegMatchWorker/SegMatch/filter_duplicate_segments: True
 * /SegMapper/SegMatchWorker/SegMatch/min_time_between_segment_for_matches_s: 60
 * /SegMapper/SegMatchWorker/align_target_map_on_first_loop_closure: False
 * /SegMapper/SegMatchWorker/close_loops: True
 * /SegMapper/SegMatchWorker/distance_between_segmentations_m: 0.0
 * /SegMapper/SegMatchWorker/distance_to_lower_target_cloud_for_viz_m: 15.0
 * /SegMapper/SegMatchWorker/export_segments_and_matches: False
 * /SegMapper/SegMatchWorker/line_scale_loop_closures: 5.0
 * /SegMapper/SegMatchWorker/line_scale_matches: 0.5
 * /SegMapper/SegMatchWorker/localize: False
 * /SegMapper/SegMatchWorker/publish_predicted_segment_matches: False
 * /SegMapper/SegMatchWorker/ratio_of_points_to_keep_when_publishing: 1.0
 * /SegMapper/clear_local_map_after_loop_closure: False
 * /SegMapper/number_of_robots: 1
 * /SegMapper/publish_world_to_odom: True
 * /SegMapper/robot_prefix: na
 * /SegMapper/tf_publication_rate_hz: 10
 * /SegMapper/world_frame: map
 * /rosdistro: kinetic
 * /rosversion: 1.12.14
 * /segmapper/icp_configuration_file: /home/lee/segmap_...
 * /segmapper/icp_input_filters_file: /home/lee/segmap_...
 * /use_sim_time: True

NODES
  /
    segmapper (segmapper/segmapper_node)
    visualizer (rviz/rviz)

auto-starting new master
process[master]: started with pid [5650]
ROS_MASTER_URI=http://localhost:11311

setting /run_id to a4868456-be69-11e9-9c48-441ca8e3ae23
process[rosout-1]: started with pid [5663]
started core service [/rosout]
process[visualizer-2]: started with pid [5687]
process[segmapper-3]: started with pid [5688]
GNU gdb (Ubuntu 7.11.1-0ubuntu1~16.5) 7.11.1
Copyright (C) 2016 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.  Type "show copying"
and "show warranty" for details.
This GDB was configured as "x86_64-linux-gnu".
Type "show configuration" for configuration details.
For bug reporting instructions, please see:
<http://www.gnu.org/software/gdb/bugs/>.
Find the GDB manual and other documentation resources online at:
<http://www.gnu.org/software/gdb/documentation/>.
For help, type "help".
Type "apropos word" to search for commands related to "word"...
Reading symbols from /home/lee/segmap_ws/devel/lib/segmapper/segmapper_node...[ INFO] [1565769656.321547658]: rviz version 1.12.17
[ INFO] [1565769656.321587837]: compiled against Qt version 5.5.1
[ INFO] [1565769656.321598683]: compiled against OGRE version 1.9.0 (Ghadamon)
done.
[ INFO] [1565769656.891955334]: Stereo is NOT SUPPORTED
[ INFO] [1565769656.892026226]: OpenGl version: 4.6 (GLSL 4.6).
0x1487320 void QWindowPrivate::setTopLevelScreen(QScreen*, bool) ( QScreen(0x893540) ): Attempt to set a screen on a child window.
0x1477200 void QWindowPrivate::setTopLevelScreen(QScreen*, bool) ( QScreen(0x893540) ): Attempt to set a screen on a child window.
0x1479bf0 void QWindowPrivate::setTopLevelScreen(QScreen*, bool) ( QScreen(0x893540) ): Attempt to set a screen on a child window.
0x1486970 void QWindowPrivate::setTopLevelScreen(QScreen*, bool) ( QScreen(0x893540) ): Attempt to set a screen on a child window.
Starting program: /home/lee/segmap_ws/devel/lib/segmapper/segmapper_node __name:=segmapper __log:=/home/lee/.ros/log/a4868456-be69-11e9-9c48-441ca8e3ae23/segmapper-3.log
[tcsetpgrp failed in terminal_inferior: 对设备不适当的 ioctl 操作]
[tcsetpgrp failed in terminal_inferior: 对设备不适当的 ioctl 操作]
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1".
[tcsetpgrp failed in terminal_inferior: 对设备不适当的 ioctl 操作]
[New Thread 0x7fff889ed700 (LWP 5859)]
[New Thread 0x7fff83fff700 (LWP 5860)]
[New Thread 0x7fff837fe700 (LWP 5861)]
[New Thread 0x7fff82ffd700 (LWP 5873)]
WARNING: Logging before InitGoogleLogging() is written to STDERR
I0814 16:02:00.525790  5794 segmatch.cpp:23] Do not forget to initialize SegMatch.
I0814 16:02:00.593253  5794 laser_track.cpp:16] Loading ICP configurations from: /home/lee/segmap_ws/src/segmap/segmapper/launch/kitti/icp_dynamic_outdoor.yaml
I0814 16:02:00.599375  5794 laser_track.cpp:26] Loading ICP input filters from: /home/lee/segmap_ws/src/segmap/segmapper/launch/kitti/input_filters_outdoor.yaml
I0814 16:02:00.601919  5794 incremental_estimator.cpp:31] Creating loop closure noise model with cauchy.
I0814 16:02:00.602021  5794 incremental_estimator.cpp:54] Loading ICP configurations from: /home/lee/segmap_ws/src/segmap/segmapper/launch/kitti/icp_dynamic_outdoor.yaml
[ INFO] [1565769720.607272736]: Number of laser_slam workers: 1
I0814 16:02:00.608141  5794 segmapper.cpp:86] Robot 0 subscribes to /velodyne_points world and velodyne
I0814 16:02:00.608188  5794 segmapper.cpp:89] Robot 0 publishes to trajectory and local_map
[New Thread 0x7fff827fc700 (LWP 6001)]
[New Thread 0x7fff81ffb700 (LWP 6029)]
I0814 16:02:00.713829  6029 segmapper.cpp:121] publishing local maps
[New Thread 0x7fff817fa700 (LWP 6030)]
[New Thread 0x7fff80ff9700 (LWP 6031)]
[ WARN] [1565769770.322090513, 1317376479.465850308]: [SegMapper] Timeout while waiting between world and velodyne.
I0814 16:02:50.423069  6029 segmapper.cpp:121] publishing local maps
I0814 16:02:50.423118  6029 segmapper.cpp:121] publishing local maps
[New Thread 0x7fff63fff700 (LWP 6055)]
[New Thread 0x7fff637fe700 (LWP 6056)]
[New Thread 0x7fff62ffd700 (LWP 6057)]
I0814 16:02:50.765403  6029 segmapper.cpp:121] publishing local maps
I0814 16:02:51.098009  6029 segmapper.cpp:121] publishing local maps
I0814 16:02:51.430085  6029 segmapper.cpp:121] publishing local maps
I0814 16:02:51.768396  6029 segmapper.cpp:121] publishing local maps
I0814 16:02:52.100013  6029 segmapper.cpp:121] publishing local maps
I0814 16:02:52.423727  6029 segmapper.cpp:121] publishing local maps
I0814 16:02:52.756546  6029 segmapper.cpp:121] publishing local maps
I0814 16:02:53.098218  6029 segmapper.cpp:121] publishing local maps
I0814 16:02:53.434726  6029 segmapper.cpp:121] publishing local maps
I0814 16:02:53.758296  6029 segmapper.cpp:121] publishing local maps
I0814 16:02:54.090392  6029 segmapper.cpp:121] publishing local maps
I0814 16:02:54.425179  6029 segmapper.cpp:121] publishing local maps
I0814 16:02:54.765620  6029 segmapper.cpp:121] publishing local maps
I0814 16:02:55.102035  6029 segmapper.cpp:121] publishing local maps
I0814 16:02:55.423818  6029 segmapper.cpp:121] publishing local maps
I0814 16:02:55.758056  6029 segmapper.cpp:121] publishing local maps
I0814 16:02:56.090567  6029 segmapper.cpp:121] publishing local maps
I0814 16:02:56.433516  6029 segmapper.cpp:121] publishing local maps
I0814 16:02:56.771297  6029 segmapper.cpp:121] publishing local maps
I0814 16:02:57.091809  6029 segmapper.cpp:121] publishing local maps
I0814 16:02:57.423559  6029 segmapper.cpp:121] publishing local maps
I0814 16:02:57.766336  6029 segmapper.cpp:121] publishing local maps
I0814 16:02:58.100080  6029 segmapper.cpp:121] publishing local maps
I0814 16:02:58.437070  6029 segmapper.cpp:121] publishing local maps
I0814 16:02:58.758193  6029 segmapper.cpp:121] publishing local maps
I0814 16:02:59.092109  6029 segmapper.cpp:121] publishing local maps
I0814 16:02:59.423753  6029 segmapper.cpp:121] publishing local maps
I0814 16:02:59.757546  6029 segmapper.cpp:121] publishing local maps
I0814 16:03:00.103891  6029 segmapper.cpp:121] publishing local maps
I0814 16:03:00.424468  6029 segmapper.cpp:121] publishing local maps
I0814 16:03:00.757823  6029 segmapper.cpp:121] publishing local maps
I0814 16:03:01.090870  6029 segmapper.cpp:121] publishing local maps
I0814 16:03:01.433143  6029 segmapper.cpp:121] publishing local maps
I0814 16:03:01.757112  6029 segmapper.cpp:121] publishing local maps
I0814 16:03:02.092766  6029 segmapper.cpp:121] publishing local maps

Thread 1 "segmapper_node" received signal SIGSEGV, Segmentation fault.
0x00007ffff5e09de8 in Eigen::numext::abs<float> (x=@0x1ffffd2002754: <error reading variable>) at /home/lee/segmap_ws/devel/include/eigen3/Eigen/src/Core/MathFunctions.h:1263

The error occurred in Eigen source code


(gdb) l
1288    }
1289    
1290    template <> EIGEN_DEVICE_FUNC EIGEN_ALWAYS_INLINE
1291    double abs(const std::complex<double>& x) {
1292      return ::hypot(x.real(), x.imag());
1293    }
1294    #endif
1295    
1296    template<typename T>
1297    EIGEN_DEVICE_FUNC EIGEN_ALWAYS_INLINE
(gdb) 
1298    T exp(const T &x) {
1299      EIGEN_USING_STD_MATH(exp);
1300      return exp(x);
1301    }
1302    
1303    #if defined(__SYCL_DEVICE_ONLY__)
1304    SYCL_SPECIALIZE_FLOATING_TYPES_UNARY(exp, exp)
1305    #endif // defined(__SYCL_DEVICE_ONLY__)
1306    
1307    #if defined(EIGEN_GPUCC)
(gdb) 
1308    template<> EIGEN_DEVICE_FUNC EIGEN_ALWAYS_INLINE
1309    float exp(const float &x) { return ::expf(x); }
1310    
1311    template<> EIGEN_DEVICE_FUNC EIGEN_ALWAYS_INLINE
1312    double exp(const double &x) { return ::exp(x); }
1313    
1314    template<> EIGEN_DEVICE_FUNC EIGEN_ALWAYS_INLINE
1315    std::complex<float> exp(const std::complex<float>& x) {
1316      float com = ::expf(x.real());
1317      float res_real = com * ::cosf(x.imag());
(gdb) 
1318      float res_imag = com * ::sinf(x.imag());
1319      return std::complex<float>(res_real, res_imag);
1320    }
1321    
1322    template<> EIGEN_DEVICE_FUNC EIGEN_ALWAYS_INLINE
1323    std::complex<double> exp(const std::complex<double>& x) {
1324      double com = ::exp(x.real());
1325      double res_real = com * ::cos(x.imag());
1326      double res_imag = com * ::sin(x.imag());
1327      return std::complex<double>(res_real, res_imag);

The backtrace also points to Eigen library:


(gdb) bt
#0  0x00007ffff5e09de8 in Eigen::numext::abs<float> (x=@0x1ffffd2002754: <error reading variable>)
    at /home/lee/segmap_ws/devel/include/eigen3/Eigen/src/Core/MathFunctions.h:1263
#1  Eigen::internal::scalar_abs_op<float>::operator() (this=0x7fffffff9940, a=@0x1ffffd2002754: <error reading variable>)
    at /home/lee/segmap_ws/devel/include/eigen3/Eigen/src/Core/functors/UnaryFunctors.h:44
#2  0x00007ffff44385ad in Eigen::internal::unary_evaluator<Eigen::CwiseUnaryOp<Eigen::internal::scalar_score_coeff_op<float>, Eigen::Block<Eigen::Matrix<float, -1, -1, 0, -1, -1>, -1, -1, false> const>, Eigen::internal::IndexBased, float>::coeff (this=0x7fffffff9940, row=0, col=1) at /usr/local/include/eigen3/Eigen/src/Core/CoreEvaluators.h:543
#3  0x00007ffff44297af in Eigen::internal::visitor_evaluator<Eigen::CwiseUnaryOp<Eigen::internal::scalar_score_coeff_op<float>, Eigen::Block<Eigen::Matrix<float, -1, -1, 0, -1, -1>, -1, -1, false> const> >::coeff (this=0x7fffffff9940, row=0, col=1) at /usr/local/include/eigen3/Eigen/src/Core/Visitor.h:79
#4  0x00007ffff44168ad in Eigen::internal::visitor_impl<Eigen::internal::max_coeff_visitor<Eigen::CwiseUnaryOp<Eigen::internal::scalar_score_coeff_op<float>, Eigen::Block<Eigen::Matrix<float, -1, -1, 0, -1, -1>, -1, -1, false> const> >, Eigen::internal::visitor_evaluator<Eigen::CwiseUnaryOp<Eigen::internal::scalar_score_coeff_op<float>, Eigen::Block<Eigen::Matrix<float, -1, -1, 0, -1, -1>, -1, -1, false> const> >, -1>::run (mat=..., visitor=...) at /usr/local/include/eigen3/Eigen/src/Core/Visitor.h:54
#5  0x00007ffff4406fcd in Eigen::DenseBase<Eigen::CwiseUnaryOp<Eigen::internal::scalar_score_coeff_op<float>, Eigen::Block<Eigen::Matrix<float, -1, -1, 0, -1, -1>, -1, -1, false> const> >::visit<Eigen::internal::max_coeff_visitor<Eigen::CwiseUnaryOp<Eigen::internal::scalar_score_coeff_op<float>, Eigen::Block<Eigen::Matrix<float, -1, -1, 0, -1, -1>, -1, -1, false> const> > > (this=0x7fffffff9b40, visitor=...) at /usr/local/include/eigen3/Eigen/src/Core/Visitor.h:116
#6  0x00007ffff43fa80a in Eigen::DenseBase<Eigen::CwiseUnaryOp<Eigen::internal::scalar_score_coeff_op<float>, Eigen::Block<Eigen::Matrix<float, -1, -1, 0, -1, -1>, -1, -1, false> const> >::maxCoeff<long> (this=0x7fffffff9b40, rowPtr=0x7fffffff9a18, colPtr=0x7fffffff9a20) at /usr/local/include/eigen3/Eigen/src/Core/Visitor.h:247
#7  0x00007ffff43ee7a5 in Eigen::FullPivHouseholderQR<Eigen::Matrix<float, -1, -1, 0, -1, -1> >::computeInPlace (this=0x7fffffff9ec0)
    at /usr/local/include/eigen3/Eigen/src/QR/FullPivHouseholderQR.h:490
#8  0x00007ffff43e0600 in Eigen::FullPivHouseholderQR<Eigen::Matrix<float, -1, -1, 0, -1, -1> >::compute<Eigen::Matrix<float, -1, -1, 0, -1, -1> > (this=0x7fffffff9ec0, 
    matrix=...) at /usr/local/include/eigen3/Eigen/src/QR/FullPivHouseholderQR.h:452
#9  0x00007ffff43d5e05 in Eigen::FullPivHouseholderQR<Eigen::Matrix<float, -1, -1, 0, -1, -1> >::FullPivHouseholderQR<Eigen::Matrix<float, -1, -1, 0, -1, -1> > (
    this=0x7fffffff9ec0, matrix=...) at /usr/local/include/eigen3/Eigen/src/QR/FullPivHouseholderQR.h:136
#10 0x00007ffff43cc9e0 in Eigen::MatrixBase<Eigen::Matrix<float, -1, -1, 0, -1, -1> >::fullPivHouseholderQr (this=0x7fffffff9d80)
    at /usr/local/include/eigen3/Eigen/src/QR/FullPivHouseholderQR.h:671
#11 0x00007ffff439d471 in DataPointsFiltersImpl<float>::SurfaceNormalDataPointsFilter::inPlaceFilter (this=0x1539040, cloud=...)
    at /home/lee/segmap_ws/src/libpointmatcher/pointmatcher/DataPointsFiltersImpl.cpp:555
#12 0x00007ffff431ff57 in PointMatcher<float>::DataPointsFilters::apply (this=0x152a8d8, cloud=...)
    at /home/lee/segmap_ws/src/libpointmatcher/pointmatcher/DataPointsFilter.cpp:122
#13 0x00007ffff4dea6e8 in laser_slam::LaserTrack::processPoseAndLaserScan (this=0x152a630, pose=..., in_scan=..., newFactors=0x7fffffffa9e0, newValues=0x7fffffffaa20, 
    is_prior=0x7fffffffa9b9) at /home/lee/segmap_ws/src/laser_slam/laser_slam/src/laser_track.cpp:146
#14 0x00007ffff7aadc68 in laser_slam_ros::LaserSlamWorker::scanCallback (this=0x152aa30, cloud_msg_in=...)
    at /home/lee/segmap_ws/src/laser_slam/laser_slam_ros/src/laser_slam_worker.cpp:134
#15 0x00007ffff7b0b1f8 in boost::_mfi::mf1<void, laser_slam_ros::LaserSlamWorker, sensor_msgs::PointCloud2_<std::allocator<void> > const&>::operator() (this=0x1533f40, 
    p=0x152aa30, a1=...) at /usr/include/boost/bind/mem_fn_template.hpp:165
#16 0x00007ffff7b03c94 in boost::_bi::list2<boost::_bi::value<laser_slam_ros::LaserSlamWorker*>, boost::arg<1> >::operator()<boost::_mfi::mf1<void, laser_slam_ros::LaserSlamWorker, sensor_msgs::PointCloud2_<std::allocator<void> > const&>, boost::_bi::list1<sensor_msgs::PointCloud2_<std::allocator<void> > const&> > (this=0x1533f50, f=..., a=...)
    at /usr/include/boost/bind/bind.hpp:313
#17 0x00007ffff7afae35 in boost::_bi::bind_t<void, boost::_mfi::mf1<void, laser_slam_ros::LaserSlamWorker, sensor_msgs::PointCloud2_<std::allocator<void> > const&>, boost::_bi::list2<boost::_bi::value<laser_slam_ros::LaserSlamWorker*>, boost::arg<1> > >::operator()<sensor_msgs::PointCloud2_<std::allocator<void> > const&> (this=0x1533f40, a1=...)
    at /usr/include/boost/bind/bind.hpp:905
#18 0x00007ffff7af187f in boost::detail::function::void_function_obj_invoker1<boost::_bi::bind_t<void, boost::_mfi::mf1<void, laser_slam_ros::LaserSlamWorker, sensor_msgs::PointCloud2_<std::allocator<void> > const&>, boost::_bi::list2<boost::_bi::value<laser_slam_ros::LaserSlamWorker*>, boost::arg<1> > >, void, sensor_msgs::PointCloud2_<std::allocator<void> > const&>::invoke (function_obj_ptr=..., a0=...) at /usr/include/boost/function/function_template.hpp:159
#19 0x00007ffff7b2c09b in boost::function1<void, sensor_msgs::PointCloud2_<std::allocator<void> > const&>::operator() (this=0x1533f38, a0=...)
    at /usr/include/boost/function/function_template.hpp:773
#20 0x00007ffff7b294f3 in ros::SubscriptionCallbackHelperT<sensor_msgs::PointCloud2_<std::allocator<void> > const&, void>::call (this=0x1533f30, params=...)
    at /opt/ros/kinetic/include/ros/subscription_callback_helper.h:144
#21 0x00007fffd4545d8d in ros::SubscriptionQueue::call() () from /opt/ros/kinetic/lib/libroscpp.so
#22 0x00007fffd44eb838 in ros::CallbackQueue::callOneCB(ros::CallbackQueue::TLS*) () from /opt/ros/kinetic/lib/libroscpp.so
#23 0x00007fffd44ed23b in ros::CallbackQueue::callAvailable(ros::WallDuration) () from /opt/ros/kinetic/lib/libroscpp.so
#24 0x00007fffd4549e39 in ros::SingleThreadedSpinner::spin(ros::CallbackQueue*) () from /opt/ros/kinetic/lib/libroscpp.so
#25 0x00007fffd452ee9b in ros::spin() () from /opt/ros/kinetic/lib/libroscpp.so
#26 0x00000000005a7727 in main (argc=1, argv=0x7fffffffd4e8) at /home/lee/segmap_ws/src/segmap/segmapper/src/segmapper_node.cpp:18

It seems option 3 cannot solve the issue. I'd like to have a try with option 1. Could you please explain more detail on how to disable MKL? I see tensorflow and eigen lib used in this project supported by MKL. Which one should I disable and rebuild? Looking forward to your reply!

Anybody there? Is this project still under the construction? It seems like there is no one solving issues and discussing problems.

I read the code roughly, and found the problem occurred when declare a dynamic matrix and assign it, such as:


switch (it->datatype)
{
    case PF::INT8: { view(dim, ptId) = T(*reinterpret_cast<const int8_t*>(fPtr)); fPtr += 1; break;}
    case PF::UINT8: { view(dim, ptId) = T(*reinterpret_cast<const uint8_t*>(fPtr)); fPtr += 1; break;}
    case PF::INT16: { view(dim, ptId) = T(*reinterpret_cast<const int16_t*>(fPtr)); fPtr += 2; break;}
    case PF::UINT16: { view(dim, ptId) = T(*reinterpret_cast<const uint16_t*>(fPtr)); fPtr += 2; break;}
    case PF::INT32: { view(dim, ptId) = T(*reinterpret_cast<const int32_t*>(fPtr)); fPtr += 4; break;}
    case PF::UINT32: { view(dim, ptId) = T(*reinterpret_cast<const uint32_t*>(fPtr)); fPtr += 4; break;}
    case PF::FLOAT32: { view(dim, ptId) = T(*reinterpret_cast<const float*>(fPtr)); fPtr += 4; break;}
    case PF::FLOAT64: {view(dim, ptId) = T(*reinterpret_cast<const double*>(fPtr)); fPtr += 8; break;}
    default: abort();
}

But as far as I know, if the declaration of a dynamic matrix with known size or using copy construction with a fixed-size matrix, just like what the code write:

View view(
     (it->name == "normal_x") ? cloud.getDescriptorRowViewByName("normals", 0) : ((it->name == "normal_y") ? cloud.getDescriptorRowViewByName("normals", 1) : ((it->name == "normal_z") ? cloud.getDescriptorRowViewByName("normals", 2) : ((isFeature[fieldId]) ? cloud.getFeatureViewByName(it->name) : cloud.getDescriptorViewByName(it->name)))) );

the assignment of the matrix should be legal. Is this the problem of the version of Eigen?

My issue was solved by using Eigen 3.2.10. At the same time, the same version of Eigen should be used in libnano, libpointmatcher and other packages.

-------------update------------------------------------- Eigen 3.3.4 is a better choice, since too old version of Eigen may be incompatible with gtsam

@smauq @tuandle @LiShuaixin sry to bother you,i really confused about how to install it.Could you please tell me how to install segmatch but not segmap?i dont have a nivid GPU,i cant find there no segmatch branch,and i install it forked by others years ago,but compile failed with laser_slam.someone tell to choose fix/pointers branch,but i cannot find it.thanks a lot

I am sorry for writing here. I followed segmap/readme.md and this issue108 to compile segmapper. In short I have the following setup:

Ubuntu 16.04 tf-gpu-1.8.0 Python 3.5.2 Compiler GCC 5.4 Bazel 0.10.0 cuDNN 7 CUDA 9 GPU Drivers 390

I used this command for tensorflow: bazel build --config=opt --config=cuda --define framework_shared_object=false tensorflow:libtensorflow_cc.so

and installed using: pip install tensorflow-gpu==1.8.0 # not from wheel

I built the packages using:

catkin build tensorflow_ros_cpp --cmake-args -DFORCE_TF_PIP_SEARCH="OFF" -DFORCE_TF_BAZEL_SEARCH="ON" -DTF_BAZEL_LIBRARY="/home/mr/tensorflow/bazel-bin/tensorflow/libtensorflow_cc.so" -DTF_BAZEL_SRC_DIR="/home/mr/tensorflow" -DTF_PYTHON_VERSION="Python 3.5" -DTF_PIP_PATH="$HOME/segmappyenv/lib/python/site-packages/tensorflow" -DTF_PIP_EXECUTABLE="$HOME/segmappyenv/bin/pip" -DTF_PYTHON_LIBRARY="/usr/lib/x86_64-linux-gnu/libpython3.5m.so.1.0"

catkin build segmapper --cmake-args -DFORCE_TF_PIP_SEARCH="OFF" -DFORCE_TF_BAZEL_SEARCH="ON" -DTF_BAZEL_LIBRARY="/home/mr/tensorflow/bazel-bin/tensorflow/libtensorflow_cc.so" -DTF_BAZEL_SRC_DIR="/home/mr/tensorflow" -DTF_PYTHON_VERSION="Python 3.5" -DTF_PIP_PATH="$HOME/segmappyenv/lib/python/site-packages/tensorflow" -DTF_PIP_EXECUTABLE="$HOME/segmappyenv/bin/pip" -DTF_PYTHON_LIBRARY="/usr/lib/x86_64-linux-gnu/libpython3.5m.so.1.0"

But during the compilation got these errors: Failed << glog_catkin:make [ Exited with code 2 ] Failed <<< glog_catkin [ 4.1 seconds ] Abandoned <<< eigen_checks [ Unrelated job failed ] Abandoned <<< libpointmatcher_ros [ Unrelated job failed ] Abandoned <<< minkindr [ Unrelated job failed ] Abandoned <<< minkindr_gtsam [ Unrelated job failed ] Abandoned <<< mincurves [ Unrelated job failed ] Abandoned <<< laser_slam [ Unrelated job failed ] Abandoned <<< laser_slam_ros [ Unrelated job failed ] Abandoned <<< tf_graph_executor [ Unrelated job failed ] Abandoned <<< segmatch [ Unrelated job failed ] Abandoned <<< segmatch_ros [ Unrelated job failed ] Abandoned <<< segmapper [ Unrelated job failed ]

Can you help ?

Did you solve that problem?

Thanks for your time knowing about my problems and please give me some help about how to solve it.

I was trying to compile segmap package. The environment is Ubuntu16.04 + cuda9.0 + cudnn7.0 + nccl 2.2. I set up tensorflow 1.8.0 by running whl file, which is by compliing form source.

bazel build --config=opt --config=cuda --cxxopt="-D_GLIBCXX_USE_CXX11_ABI=0" --define framework_shared_object=true\ //tensorflow/tools/pip_package:build_pip_package //tensorflow:libtensorflow_cc.so \ //tensorflow:libtensorflow_framework.so \ //tensorflow:install_headers

I followed up tutorial to build the tensorflow_ros_cpp first and I got its done, then i got stuck by compiling segmapper. Here is my code and reported error. catkin build tensorflow_ros_cpp --cmake-args -DFORCE_TF_PIP_SEARCH="ON" -DFORCE_TF_BAZEL_SEARCH="ON" -DFORCE_TF_CATKIN_SEARCH="ON" -DTF_BAZEL_LIBRARY="/home/haotian/tensorflow/tensorflow-1.8.0/bazel-bin/tensorflow/libtensorflow_cc.so" -DTF_BAZEL_SRC_DIR="/home/haotian/tensorflow/tensorflow-1.8.0" -DTF_PYTHON_VERSION="2.7.15" -DTF_PIP_PATH="$HOME/segmappyenv/lib/python2.7/site-packages/tensorflow" -DTF_PYTHON_LIBRARY="/usr/lib/x86_64-linux-gnu/libpython2.7.so.1.0"

catkin build segmapper --cmake-args -DFORCE_TF_PIP_SEARCH="ON" -DFORCE_TF_BAZEL_SEARCH="ON" -DFORCE_TF_CATKIN_SEARCH="ON" -DTF_BAZEL_LIBRARY="/home/haotian/tensorflow/tensorflow-1.8.0/bazel-bin/tensorflow/libtensorflow_cc.so" -DTF_BAZEL_SRC_DIR="/home/haotian/tensorflow/tensorflow-1.8.0" -DTF_PYTHON_VERSION="2.7.15" -DTF_PIP_PATH="$HOME/segmappyenv/lib/python2.7/site-packages/tensorflow" -DTF_PYTHON_LIBRARY="/usr/lib/x86_64-linux-gnu/libpython2.7.so.1.0"

reported error

/usr/bin/ld: warning: libiomp5.so, needed by /home/haotian/segmap_ws/devel/lib/libtensorflow_cc.so, not found (try using -rpath or -rpath-link) /usr/bin/ld: warning: libmklml_intel.so, needed by /home/haotian/segmap_ws/devel/lib/libtensorflow_cc.so, not found (try using -rpath or -rpath-link) /home/haotian/segmap_ws/devel/lib/libtensorflow_cc.so: undefined reference to dnnLayoutGetMemorySize_F32' /home/haotian/segmap_ws/devel/lib/libtensorflow_cc.so: undefined reference toomp_in_parallel@VERSION' /home/haotian/segmap_ws/devel/lib/libtensorflow_cc.so: undefined reference to GOMP_barrier@VERSION' /home/haotian/segmap_ws/devel/lib/libtensorflow_cc.so: undefined reference todnnLayoutCompare_F32' /home/haotian/segmap_ws/devel/lib/libtensorflow_cc.so: undefined reference to omp_get_max_threads@VERSION' /home/haotian/segmap_ws/devel/lib/libtensorflow_cc.so: undefined reference todnnConversionCreate_F32' /home/haotian/segmap_ws/devel/lib/libtensorflow_cc.so: undefined reference to omp_get_num_threads@VERSION' /home/haotian/segmap_ws/devel/lib/libtensorflow_cc.so: undefined reference todnnLayoutCreateFromPrimitive_F32' /home/haotian/segmap_ws/devel/lib/libtensorflow_cc.so: undefined reference to dnnLayoutDeserialize_F32' /home/haotian/segmap_ws/devel/lib/libtensorflow_cc.so: undefined reference todnnConversionExecute_F32' /home/haotian/segmap_ws/devel/lib/libtensorflow_cc.so: undefined reference to cblas_sgemm_batch' /home/haotian/segmap_ws/devel/lib/libtensorflow_cc.so: undefined reference toMKL_Domatcopy' /home/haotian/segmap_ws/devel/lib/libtensorflow_cc.so: undefined reference to dnnLayoutCreate_F32' /home/haotian/segmap_ws/devel/lib/libtensorflow_cc.so: undefined reference toMKL_Comatcopy' /home/haotian/segmap_ws/devel/lib/libtensorflow_cc.so: undefined reference to dnnLayoutSerializationBufferSize_F32' /home/haotian/segmap_ws/devel/lib/libtensorflow_cc.so: undefined reference tocblas_cgemm_batch' /home/haotian/segmap_ws/devel/lib/libtensorflow_cc.so: undefined reference to GOMP_parallel@VERSION' /home/haotian/segmap_ws/devel/lib/libtensorflow_cc.so: undefined reference toMKL_Somatcopy' /home/haotian/segmap_ws/devel/lib/libtensorflow_cc.so: undefined reference to dnnLayoutDelete_F32' /home/haotian/segmap_ws/devel/lib/libtensorflow_cc.so: undefined reference todnnConvolutionCreateBackwardBias_F32' /home/haotian/segmap_ws/devel/lib/libtensorflow_cc.so: undefined reference to dnnDelete_F32' /home/haotian/segmap_ws/devel/lib/libtensorflow_cc.so: undefined reference tocblas_zgemm_batch' /home/haotian/segmap_ws/devel/lib/libtensorflow_cc.so: undefined reference to dnnExecute_F32' /home/haotian/segmap_ws/devel/lib/libtf_graph_executor.so: undefined reference totensorflow::ReadBinaryProto(tensorflow::Env, std::__cxx11::basic_string<char, std::char_traits, std::allocator > const&, google::protobuf::MessageLite)' /home/haotian/segmap_ws/devel/lib/libtensorflow_cc.so: undefined reference to cblas_dgemm_batch' /home/haotian/segmap_ws/devel/lib/libtensorflow_cc.so: undefined reference toomp_get_thread_num@VERSION' /home/haotian/segmap_ws/devel/lib/libtensorflow_cc.so: undefined reference to MKL_Zomatcopy' /home/haotian/segmap_ws/devel/lib/libtensorflow_cc.so: undefined reference todnnLayoutSerialize_F32' /home/haotian/segmap_ws/devel/lib/libtf_graph_executor.so: undefined reference to `tensorflow::internal::CheckOpMessageBuilder::NewString[abi:cxx11]()' collect2: error: ld returned 1 exit status make[2]: [/home/haotian/segmap_ws/devel/lib/segmapper/segmapper_node] Error 1 make[1]: [CMakeFiles/segmapper_node.dir/all] Error 2 make: *** [all] Error 2 cd /home/haotian/segmap_ws/build/segmapper; catkin build --get-env segmapper | catkin env -si /usr/bin/make --jobserver-fds=6,7 -j; cd - ........................................................................................................................................................................................................... Failed << segmapper:make [ Exited with code 2 ]
Failed <<< segmapper [ 59.1 seconds ]
[build] Summary: 21 of 22 packages succeeded.
[build] Ignored: 9 packages were skipped or are blacklisted.
[build] Warnings: 12 packages succeeded with warnings.
[build] Abandoned: None.
[build] Failed: 1 packages failed.
[build] Runtime: 37 minutes and 34.7 seconds total.

Please give me some guidance to slove this problem. Thank you so much.

@smauq Hello, I encounter the same question with @tuandle . I built the package in a virtualenv, I used Ubuntu16.04+cuda9.0+cudnn7.0+bazel 0.10.0 and built tensorflow-r1.8 from source. When I compiled the segmap_ws package in the same virtualenv , I encounter error when I use "$ catkin build segmapper"

Errors << segmapper:make /home/james/segmap_ws/logs/segmapper/build.make.006.log

/home/james/segmap_ws/devel/lib/libtf_graph_executor.so: undefined reference to tensorflow::ReadBinaryProto(tensorflow::Env*, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&, google::protobuf::MessageLite*)' /home/james/segmap_ws/devel/lib/libtf_graph_executor.so: undefined reference totensorflow::internal::CheckOpMessageBuilder::NewStringabi:cxx11' collect2: error: ld returned 1 exit status make[2]: [/home/james/segmap_ws/devel/lib/segmapper/segmapper_node] Error 1 make[1]: [CMakeFiles/segmapper_node.dir/all] Error 2 make: *** [all] Error 2 cd /home/james/segmap_ws/build/segmapper; catkin build --get-env segmapper | catkin env -si /usr/bin/make --jobserver-fds=6,7 -j; cd - ............................................................................... Failed << segmapper:make [ Exited with code 2 ] Failed <<< segmapper [ 3.5 seconds ] [build] Summary: 21 of 22 packages succeeded. [build] Ignored: 9 packages were skipped or are blacklisted. [build] Warnings: 4 packages succeeded with warnings. [build] Abandoned: None. [build] Failed: 1 packages failed. [build] Runtime: 13.0 seconds total. I've seen your solution above,but I do not clearly understand .Could you please tell me how to deal with the problem? Thank you very much

Hi.I encountered the same problem as you, has your problem been solved?

Hi @smauq, I encounter the same problem as #106 . I compiled and pip installed the newest TF, version 1.10.

(segmappyenv) tuan@tuan:~/segmap_ws$ pip show tensorflow
Name: tensorflow
Version: 1.10.0
Summary: TensorFlow is an open source machine learning framework for everyone.
Home-page: https://www.tensorflow.org/
Author: Google Inc.
Author-email: opensource@google.com
License: Apache 2.0
Location: /home/tuan/segmappyenv/lib/python2.7/site-packages
Requires: setuptools, astor, tensorboard, enum34, gast, keras-preprocessing, six, keras-applications, wheel, absl-py, backports.weakref, termcolor, protobuf, numpy, grpcio, mock
Required-by: 
(segmappyenv) tuan@tuan:~/segmap_ws$

But I can't build segmapper. The error stays the same

__________________________________________________________________________________________________________________
Errors     << segmapper:make /home/tuan/segmap_ws/logs/segmapper/build.make.002.log                               
/home/tuan/segmap_ws/devel/lib/libtf_graph_executor.so: undefined reference to `tensorflow::ReadBinaryProto(tensorflow::Env*, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&, google::protobuf::MessageLite*)'
/home/tuan/segmap_ws/devel/lib/libtf_graph_executor.so: undefined reference to `tensorflow::internal::CheckOpMessageBuilder::NewString[abi:cxx11]()'
collect2: error: ld returned 1 exit status
make[2]: *** [/home/tuan/segmap_ws/devel/lib/segmapper/segmapper_node] Error 1
make[1]: *** [CMakeFiles/segmapper_node.dir/all] Error 2
make: *** [all] Error 2
cd /home/tuan/segmap_ws/build/segmapper; catkin build --get-env segmapper | catkin env -si  /usr/bin/make --jobserver-fds=6,7 -j; cd -
..................................................................................................................
Failed     << segmapper:make                     [ Exited with code 2 ]                                           
Failed    <<< segmapper                          [ 3.4 seconds ]                                                  
[build] Summary: 21 of 22 packages succeeded.                                                                     
[build]   Ignored:   9 packages were skipped or are blacklisted.                                                  
[build]   Warnings:  4 packages succeeded with warnings.                                                          
[build]   Abandoned: None.                                                                                        
[build]   Failed:    1 packages failed.                                                                           
[build] Runtime: 7.2 seconds total.                                                                               
(segmappyenv) tuan@tuan:~/segmap_ws$

What should I do next? Thank you for your help!

hi,Could you please tell me more about how to solve it because I encountered the same problem,thanks!

ethz-asl / segmap

can't build segmapper #108

Errors << segmapper:make /home/james/segmap_ws/logs/segmapper/build.make.006.log