Failing test (CRASH)

[pastoreg]

Hi all,

I have an MR that is failing because of a test failure (CRASH). It seems the test runs fine in serial or with mpiexec, however, ./run_tests --n-threads=2 reproduces the error, as below. I am not sure what may be causing this and would appreciate any help. Thanks!

(moose) gpastore@x86_64-apple-darwin13 bison % ./run_tests --n-threads=2 --re=test:tensor_mechanics/fecral_plasticity/power-law_strain_hardening.test_power-law_plasticity_fecral_tm

test:tensor_mechanics/fecral_plasticity/power-law_strain_hardening.test_power-law_plasticity_fecral_tm: Exit Code: -11

test:tensor_mechanics/fecral_plasticity/power-law_strain_hardening.test_power-law_plasticity_fecral_tm: ################################################################################

test:tensor_mechanics/fecral_plasticity/power-law_strain_hardening.test_power-law_plasticity_fecral_tm: Tester failed, reason: CRASH

test:tensor_mechanics/fecral_plasticity/power-law_strain_hardening.test_power-law_plasticity_fecral_tm:

test:tensor_mechanics/fecral_plasticity/power-law_strain_hardening.test_power-law_plasticity_fecral_tm [...] FAILED (CRASH)

Final Test Results:

---

test:tensor_mechanics/fecral_plasticity/power-law_strain_hardening.test_power-law_plasticity_fecral_tm [...] FAILED (CRASH)

---

Ran 1 tests in 1.5 seconds. Average test time 0.2 seconds, maximum test time 0.2 seconds.

Giovanni

--
Giovanni Pastore, PhD
Research Associate Professor
Department of Nuclear Engineering
University of Tennessee
Knoxville, TN 37916
gpastore@utk.edu

[pastoreg]

Never mind, I figured this out.

The crash was due to a segfault caused by a pretty simple error. Below is a code snippet from a function called by my constructor:

if (!_fe_problem.hasFunction("FeCrAl_hardening_function_296K"))
{
const std::vector hf0_plastic_strains = {0.0e+00,
1.0e-03,
2.0e-03,
3.0e-03,
4.0e-03,
5.0e-03,
1.0e-02,
2.0e-02,
3.0e-02,
4.0e-02,
5.0e-02,
6.0e-02,
7.0e-02,
8.0e-02,
9.0e-02,
1.0e-01,
1.1e-01,
1.6e-01};
const std::vector hf0_stresses = {4.55e+08,
4.65e+08,
4.71e+08,
4.75e+08,
4.79e+08,
4.82e+08,
4.94e+08,
5.12e+08,
5.26e+08,
5.39e+08,
5.51e+08,
5.63e+08,
5.74e+08,
5.85e+08,
5.96e+08,
6.07e+08,
6.17e+08,
6.72e+08};

InputParameters hf0_params = _app.getFactory().getValidParams("PiecewiseLinear");
hf0_params.set<std::vector<Real>>("x") = hf0_plastic_strains;
hf0_params.set<std::vector<Real>>("y") = hf0_stresses;
_fe_problem.addFunction("PiecewiseLinear", "FeCrAl_hardening_function_296K", hf0_params);

}
_hf_temperatures.push_back(296.);
_hardening_functions.push_back(dynamic_cast<const PiecewiseLinear *>(
&_fe_problem.getFunction("FeCrAl_hardening_function_296K")));

In the version that was giving the test crash (segfaulting), the last two lines were inside the if statement, hence, the _hf_temperatures and _hardening_functions ended up having size zero in multi-thread.

Thanks,
Giovanni

[aeslaughter]

Thanks for posting the solution.