Change type of nextTimerDefined from int to modelica_boolean in handlerTimersFMI

[Arinomo]

Type mismatched of argument 4 in handlersTimersFMI when integrating FMU from OM in rt-env
Discussion 11436

Purpose

So FMU from OM could be integrated in rt-env

Approach

-int handleTimersFMI(DATA* data, threadData_t *threadData, double currentTime, int *nextTimerDefined, double *nextTimerActivationTime)
+int handleTimersFMI(DATA* data, threadData_t *threadData, double currentTime, modelica_boolean *nextTimerDefined, double *nextTimerActivationTime)

[CLAassistant]

All committers have signed the CLA.

[adrpo]

I had to kill DualMassOscillator_cs.mos test as it got stuck. We should really use alarm on these tests as they seem to get stuck.

hudson    317996  0.0  0.0  12468  9932 ?        S    15:03   0:00 /usr/bin/perl -w ../../rtest -v -nolib DualMassOscillator_cs.mos
hudson    318001  0.0  0.0   2888   960 ?        S    15:03   0:00 sh -c /var/lib/jenkins/ws/OpenModelica_PR-11469/build/bin/omc --running-testsuite=/tmp/rtest16086.aSIXp.tmpDualMassOscillator_cs.mos +locale=C --alarm=900  DualMassOscillator_cs.mos >>/tmp/omc-rtest-unknown/omsimulator/DualMassOscillator_cs.mos_temp4172/log-DualMassOscillator_cs.mos 2>&1
hudson    318003  0.0  1.2 444624 397680 ?       S    15:03   0:03 /var/lib/jenkins/ws/OpenModelica_PR-11469/build/bin/omc --running-testsuite=/tmp/rtest16086.aSIXp.tmpDualMassOscillator_cs.mos +locale=C --alarm=900 DualMassOscillator_cs.mos
hudson    319087  0.0  0.0   2888   964 ?        S    15:03   0:00 /bin/sh -c /var/lib/jenkins/ws/OpenModelica_PR-11469/build/bin/../bin/OMSimulator DualMassOscillator_cs.lua
hudson    319089  1.0  8.8 2949720 2904092 ?     S    15:03   3:05 /var/lib/jenkins/ws/OpenModelica_PR-11469/build/bin/../bin/OMSimulator DualMassOscillator_cs.lua

[AnHeuermann]

A lot of strange side effects for FMUs. I guess we need to reproduce these issues locally to fix them.

[AnHeuermann]

Looks like the type cast from fmi2Boolean to modelica_boolean isn't working as expected. Since we don't use the FMI or OpenModelica types in a clean way I mixed it even further and changed

- fmi2Boolean nextTimerDefined;
+ modelica_boolean nextTimerDefined;

For FMI we should define modelica_boolean and the other types as the FMI equivalent types, see #11577

[AnHeuermann]

@Arinomo Can you check if this works for your dSPACE application? Then I'll merge the PR.

[Arinomo]

@AnHeuermann ok, i will check it and will report as soon as next week.

[Arinomo]

@AnHeuermann it works for my Application. Thank you.