Catch segmentation fault for broken FMU

[AnHeuermann]

Description

When testing Rectifier.fmu from FMI Cross-Check we get a segmentation fault. This happens in other tools as well and the FMU is marked as notCompliantWithLatestRules.

Steps to reproduce the behavior

To simulate FMU Rectifier.fmu with OMSimulator run

$ OMSimulator Rectifier.lua

Lua file:

-- Lua file for Rectifier.fmu
oms_setTempDirectory("temp")
oms_newModel("model")
oms_addSystem("model.root", oms_system_wc)

-- instantiate FMU
oms_addSubModel("model.root.fmu", "Rectifier.fmu")

-- Simulation settings
oms_setSignalFilter("model", ".*")
oms_setResultFile("model", "Rectifier_out.csv")
oms_setStartTime("model", 0.0)
oms_setStopTime("model", 0.1)
oms_setTolerance("model", 1e-8)
initialStepSize, minimumStepSize, maximumStepSize, status = oms_getVariableStepSize("model")
oms_setVariableStepSize("model", 1e-07, minimumStepSize, 1e-07)
oms_setFixedStepSize("model", 1e-07)

-- Instantiate, initialize and simulate
oms_instantiate("model")
oms_initialize("model")
oms_simulate("model")
oms_terminate("model")
oms_delete("model")

Expected behavior

If possible I would like to be able to catch this error before failing with a segmentation fault and reporting a helpful error message to the user.

Output from gdb

When running it with gdb we get:

$ gdb OMSimulator 
(gdb) r Rectifier.lua
Starting program: /home/aheuermann1/workspace/OpenModelica/build/bin/OMSimulator Rectifier.lua
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1".
info:    Set temp directory to    "/home/aheuermann1/workspace/OpenModelica/OMSimulator/testsuite/fmi-cross-check/fmi-cross-check/results/2.0/cs/linux64/OMSimulator/v2.1.0/MapleSim/2016.1/Rectifier"
info:    Set working directory to "/home/aheuermann1/workspace/OpenModelica/OMSimulator/testsuite/fmi-cross-check/fmi-cross-check/results/2.0/cs/linux64/OMSimulator/v2.1.0/MapleSim/2016.1/Rectifier"
info:    Set temp directory to    "/home/aheuermann1/workspace/OpenModelica/OMSimulator/testsuite/fmi-cross-check/fmi-cross-check/results/2.0/cs/linux64/OMSimulator/v2.1.0/MapleSim/2016.1/Rectifier/temp"
info:    New model "model" with corresponding temp directory "/home/aheuermann1/workspace/OpenModelica/OMSimulator/testsuite/fmi-cross-check/fmi-cross-check/results/2.0/cs/linux64/OMSimulator/v2.1.0/MapleSim/2016.1/Rectifier/temp/model-8yutriof"
error:   [fmiLogger] module FMI2XML: Start attribute is required for this causality, variability and initial combination

Program received signal SIGSEGV, Segmentation fault.
oms::Variable::getCausality (this=0xe2d20000086d13c0) at /home/aheuermann1/workspace/OpenModelica/OMSimulator/src/OMSimulatorLib/Variable.cpp:88
88        switch (causality)
(gdb) bt
#0  oms::Variable::getCausality (this=0xe2d20000086d13c0) at /home/aheuermann1/workspace/OpenModelica/OMSimulator/src/OMSimulatorLib/Variable.cpp:88
#1  0x0000000000568e7b in oms::Variable::makeConnector (this=0xe2d20000086d13c0) at /home/aheuermann1/workspace/OpenModelica/OMSimulator/src/OMSimulatorLib/../OMSimulatorLib/Variable.h:90
#2  oms::ComponentFMUCS::initializeDependencyGraph_initialUnknowns (this=<optimized out>) at /home/aheuermann1/workspace/OpenModelica/OMSimulator/src/OMSimulatorLib/ComponentFMUCS.cpp:380
#3  0x0000000000568227 in oms::ComponentFMUCS::NewComponent (cref=..., parentSystem=<optimized out>, fmuPath=...) at /home/aheuermann1/workspace/OpenModelica/OMSimulator/src/OMSimulatorLib/ComponentFMUCS.cpp:240
#4  0x00000000004e1e13 in oms::System::addSubModel (this=0x9f8910, cref=..., Python Exception <class 'gdb.error'> There is no member named _M_dataplus.:
path=) at /home/aheuermann1/workspace/OpenModelica/OMSimulator/src/OMSimulatorLib/System.cpp:286
#5  0x00000000004b6030 in oms_addSubModel (cref=<optimized out>, fmuPath=0x9eda58 "../../../../../../../../../fmus/2.0/cs/linux64/MapleSim/2016.1/Rectifier/Rectifier.fmu") at /home/aheuermann1/workspace/OpenModelica/OMSimulator/src/OMSimulatorLib/OMSimulator.cpp:807
#6  0x000000000049f521 in OMSimulatorLua_oms_addSubModel (L=0x9edac8) at /home/aheuermann1/workspace/OpenModelica/OMSimulator/src/OMSimulatorLib/../OMSimulatorLua/OMSimulatorLua.c:895
#7  0x000000000063fa16 in luaD_precall (L=0x9edac8, func=0x9ed1f0, nresults=0) at ldo.c:434
#8  0x000000000064b2ed in luaV_execute (L=0x9edac8) at lvm.c:1134
#9  0x000000000063fc74 in luaD_call (L=0x9edac8, func=<optimized out>, nResults=<optimized out>) at ldo.c:499
#10 luaD_callnoyield (L=0x9edac8, func=<optimized out>, nResults=<optimized out>) at ldo.c:509
#11 0x000000000063eeae in luaD_rawrunprotected (L=0x9edac8, f=0x63d3d0 <f_call>, ud=0x7fffffffd890) at ldo.c:142
#12 0x000000000064024b in luaD_pcall (L=0x9edac8, func=0x9c4010, u=0x9c4010, old_top=16, ef=0) at ldo.c:729
#13 0x000000000063d340 in lua_pcallk (L=0x9edac8, nargs=<optimized out>, nresults=0, errfunc=<optimized out>, ctx=<optimized out>, k=0x335b4341692e3172) at lapi.c:968
#14 0x00000000004bc886 in oms_RunFile (filename_=<optimized out>) at /home/aheuermann1/workspace/OpenModelica/OMSimulator/src/OMSimulatorLib/OMSimulator.cpp:1101
#15 0x000000000059b5b8 in oms::Flags::SetCommandLineOption (cmd=...) at /home/aheuermann1/workspace/OpenModelica/OMSimulator/src/OMSimulatorLib/Flags.cpp:168
#16 0x00000000004a72fc in oms_setCommandLineOption (cmd=0x7fffffffdbf8 "Rectifier.lua") at /home/aheuermann1/workspace/OpenModelica/OMSimulator/src/OMSimulatorLib/OMSimulator.cpp:373
#17 0x000000000049e12e in main (argc=<optimized out>, argv=0x7fffffffdd18) at /home/aheuermann1/workspace/OpenModelica/OMSimulator/src/OMSimulator/main.cpp:45

OMSimulator/src/OMSimulatorLib/Variable.cpp:

oms_causality_enu_t oms::Variable::getCausality() const
{
  switch (causality)
  {
  case fmi2_causality_enu_input:
    return oms_causality_input;

  case fmi2_causality_enu_output:
    return oms_causality_output;

  case fmi2_causality_enu_parameter:
    return oms_causality_parameter;

  default:
    return oms_causality_undefined;
  }
}

Version and OS

• Version: OMSimulator v2.1.0-dev-191-gc7cab4b-linux
• OS: Ubuntu 18.04

[superlou]

Any chance this is related to #767? The seg. fault seems to be a different mechanism, but maybe the root cause is similar.

[stale[bot]]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.

[stale[bot]]

Closing this issue after a prolonged period of inactivity. If this issue is still present in the latest release, please feel free to create a new issue with up-to-date information.