compiled model crashes for some integration times

[alexander_rusako@…]

Hi,

I am a bit new to open modelica. I built my model and see that it may or may not crash for some integration time.

Model with stop time 80 and 200 it works fine. But integration time 100 leads to crash. I managed to reproduce the problem with old open modelica versions as well.

Below is part of the log. I am also attaching models and runExp.mos script which runs the model.

simulate(OneMuscleWithGland, startTime = 0.0, stopTime=100)

record SimulationResult

resultFile = "",
simulationOptions = "startTime = 0.0, stopTime = 100.0, numberOfIntervals = 5000, tolerance = 1e-06, method = 'dassl', fileNamePrefix = 'OneMuscleWithGland', options = , outputFormat = 'mat', variableFilter = '.*', cflags = , simflags = ",
messages = "Simulation execution failed for model: OneMuscleWithGland

assert | debug | division by zero at time 55.45137084070143, (a=1.000010000000006e-05) / (b=0), where divisor b expression is: if m.MF <= 0.0 then 0.0 else 1.0
stdout | warning | Integrator attempt to handle a problem with a called assert.
assert | debug | division by zero at time 55.44, (a=9.999606888309255e-06) / (b=0), where divisor b expression is: if m.MF <= 0.0 then 0.0 else 1.0
stdout | info | model terminate | Simulation terminated by an assert at time: 55.44

Limited backtrace at point of segmentation fault
/lib/x86_64-linux-gnu/libpthread.so.0(+0xfcb0)[0x7fd1e3927cb0]
/lib/x86_64-linux-gnu/libc.so.6(cfree+0x1c)[0x7fd1e35dc24c]
/usr/lib/omc/libSimulationRuntimeC.so(deInitializeDataStruc+0x1d)[0x7fd1e3e86cbd]
/usr/lib/omc/libSimulationRuntimeC.so(_main_SimulationRuntime+0x154)[0x7fd1e3ebf3d4]
/tmp/OpenModelica/OneMuscleWithGland(main+0x15e)[0x405e1e]
/lib/x86_64-linux-gnu/libc.so.6(libc_start_main+0xed)[0x7fd1e357a76d]
/tmp/OpenModelica/OneMuscleWithGland[0x4033a9]
Segmentation fault (core dumped)
",

timeFrontend = 0.017031358,
timeBackend = 0.016784455,
timeSimCode = 0.015486948,
timeTemplates = 0.011841404,
timeCompile = 0.74921455

end SimulationResult;
Warning: The initial conditions are not fully specified. Use +d=initialization for more information.

[alexander_rusako@…]

mo and mos file to reproduce crash

[Martin Sjölund]

This happens only when running using method="dassl" (localData[0]->realVars becomes 0x21 or something like that instead of being a pointer to a real array).

[alexander_rusako@…]

Do you mean that workaround is to switch to another integration method?

thanks
Alexander

[Martin Sjölund]

No, I meant that wbraun should fix it because he knows the dassl code. But yes, a workaround is to switch the integration method until it is fixed.

[Martin Sjölund]

This ticket was not closed for 1.9.1, which has now been released. It was batch modified for milestone 1.9.2 (but maybe an empty milestone was more appropriate; feel free to change it).

[alexander_rusako@…]

I can not push you to fix it in specific branch/release. But i would be glad to get ticket fixed soon as seems still blocks because other than dassl methods seem to be not so robust.

thanks
Alexander

[Willi Braun]

fixed in r23223.