Seg-fault during Material destructor

[osrf-migration]

Original report (archived issue) by Steve Peters (Bitbucket: Steven Peters, GitHub: scpeters).

---

As mentioned in release-tools #20, there are some tests that seg-fault so late that Jenkins thinks that they passed. Specifically, the INTEGRATION_concave_mesh test has been seg-faulting since pull request #1066. Through testing, I have identified the tests involving submeshes (ie. --gtest_filter="*SubmeshCollisionTest*" or --gtest_filter="*SubmeshNoCollisionTest*").

Note that it is not listed in the list of failing tests, but it is shown as seg-faulting in the
console output:

145: [----------] 9 tests from PhysicsEngines/ConcaveMeshTest (36369 ms total)
145: 
145: [----------] Global test environment tear-down
145: [==========] 9 tests from 1 test case ran. (36369 ms total)
145: [  PASSED  ] 9 tests.
145: �[0m
  5/108 Test #145: INTEGRATION_concave_mesh .....................***Exception: SegFault 58.20 sec
test 146
        Start 146: check_INTEGRATION_concave_mesh

146: Test command: /var/lib/jenkins/workspace/gazebo-default-devel-precise-amd64-gpu-nvidia/gazebo/tools/check_test_ran.py "/var/lib/jenkins/workspace/gazebo-default-devel-precise-amd64-gpu-nvidia/build/test_results/INTEGRATION_concave_mesh.xml"
146: Test timeout computed to be: 9.99988e+06
146: Checking for test results in /var/lib/jenkins/workspace/gazebo-default-devel-precise-amd64-gpu-nvidia/build/test_results/INTEGRATION_concave_mesh.xml
  6/108 Test #146: check_INTEGRATION_concave_mesh ...............   Passed    0.02 sec

Here's a backtrace that implicates the Material destructor:

(gdb) bt
#0  0x00007ffff46a6f77 in __GI_raise (sig=sig@entry=6) at ../nptl/sysdeps/unix/sysv/linux/raise.c:56
#1  0x00007ffff46aa5e8 in __GI_abort () at abort.c:90
#2  0x00007ffff46e44fb in __libc_message (do_abort=do_abort@entry=2, fmt=fmt@entry=0x7ffff47f8240 "*** Error in `%s': %s: 0x%s ***\n")
    at ../sysdeps/unix/sysv/linux/libc_fatal.c:199
#3  0x00007ffff46ef4c6 in malloc_printerr (action=3, str=0x7ffff47f41e8 "corrupted double-linked list", ptr=<optimized out>) at malloc.c:4923
#4  0x00007ffff46f0cb9 in _int_free (av=0x7fffcc000020, p=0x7fffccc999e0, have_lock=0) at malloc.c:3933
#5  0x00007ffff7b87145 in ~basic_string (this=0x7fffccc998c0, __in_chrg=<optimized out>) at /usr/include/c++/4.8/bits/basic_string.h:539
#6  gazebo::common::Material::~Material (this=0x7fffccc998b0, __in_chrg=<optimized out>)
    at /home/scpeters/ws/gazebo/src/gazebo/gazebo/common/Material.cc:65
#7  0x00007ffff7b871f9 in gazebo::common::Material::~Material (this=0x7fffccc998b0, __in_chrg=<optimized out>)
    at /home/scpeters/ws/gazebo/src/gazebo/gazebo/common/Material.cc:68
#8  0x00007ffff7b87496 in gazebo::common::Mesh::~Mesh (this=0x7fffccaccaf0, __in_chrg=<optimized out>)
    at /home/scpeters/ws/gazebo/src/gazebo/gazebo/common/Mesh.cc:54
#9  0x00007ffff7b87559 in gazebo::common::Mesh::~Mesh (this=0x7fffccaccaf0, __in_chrg=<optimized out>)
    at /home/scpeters/ws/gazebo/src/gazebo/gazebo/common/Mesh.cc:59
#10 0x00007ffff7b8ba3f in gazebo::common::MeshManager::~MeshManager (
    this=0x7ffff78fb9e0 <SingletonT<gazebo::common::MeshManager>::GetInstance()::t>, __in_chrg=<optimized out>)
    at /home/scpeters/ws/gazebo/src/gazebo/gazebo/common/MeshManager.cc:79
#11 0x00007ffff46ac071 in __run_exit_handlers (status=0, listp=0x7ffff4a316a8 <__exit_funcs>, run_list_atexit=run_list_atexit@entry=true)
    at exit.c:77
#12 0x00007ffff46ac0f5 in __GI_exit (status=<optimized out>) at exit.c:99
#13 0x00007ffff4691dec in __libc_start_main (main=0x412ba0 <main(int, char**)>, argc=2, ubp_av=0x7fffffffda78, init=<optimized out>, 
    fini=<optimized out>, rtld_fini=<optimized out>, stack_end=0x7fffffffda68) at libc-start.c:294
#14 0x0000000000415ced in _start ()
(gdb)

@iche033 @chapulina

[osrf-migration]

Original comment by Ian Chen (Bitbucket: Ian Chen, GitHub: iche033).

---

pull request #1105

[osrf-migration]

Original comment by Hugo Boyer (Bitbucket: hugomatic, GitHub: hugomatic).

---

• changed state from "new" to "resolved"

Resolved in pull request #1105

[osrf-migration]

Original comment by Nate Koenig (Bitbucket: Nathan Koenig).

---

• set version to "all"

[osrf-migration]

Original comment by Nate Koenig (Bitbucket: Nathan Koenig).

---

• changed state from "resolved" to "closed"