Validation of nightly Cpp tests broken

[Rüdiger Franke]

Since yesterday the validation fails for almost all Cpp examples, see:

​https://test.openmodelica.org/libraries/MSL_3.2.1_cpp/BuildModelRecursive.html

It looks like some variables are not obtained anymore from the results file, like the drum temperature T_S that is between 100 and 300 degC, while the validation obtains a constant 1.80e7, see:

​https://test.openmodelica.org/libraries/MSL_3.2.1_cpp/files/Modelica.Fluid.Examples.DrumBoiler.DrumBoiler.diff.T_S.html

Hudson's regression tests work with validation. Also my local build works.

What is wrong with the nightly tests?

[Martin Sjölund]

You must have changed something in the cppruntime because the output of the csv-file is actually (for BusUsage):

"time","gain.u","gain.y","part.realExpression.y","controlBus.realSignal1","controlBus.subControlBus.myRealSignal","part.subControlBus.myRealSignal","sine.y","integerStep.y","controlBus.integerSignal","booleanStep.y","part.booleanExpression.y","controlBus.booleanSignal","controlBus.subControlBus.myBooleanSignal","part.subControlBus.myBooleanSignal",
0,0x225be00,0x225be08,0x225be10,0x225be00,0x225be10,0x225be10,0x225be00,0x225bd80,0x225bd80,0x224cc40,0x224cc41,0x224cc40,0x224cc41,0x224cc41,
0.004,0x225be00,0x225be08,0x225be10,0x225be00,0x225be10,0x225be10,0x225be00,0x225bd80,0x225bd80,0x224cc40,0x224cc41,0x224cc40,0x224cc41,0x224cc41,
0.008,0x225be00,0x225be08,0x225be10,0x225be00,0x225be10,0x225be10,0x225be00,0x225bd80,0x225bd80,0x224cc40,0x224cc41,0x224cc40,0x224cc41,0x224cc41,
0.012,0x225be00,0x225be08,0x225be10,0x225be00,0x225be10,0x225be10,0x225be00,0x225bd80,0x225bd80,0x224cc40,0x224cc41,0x224cc40,0x224cc41,0x224cc41,
0.016,0x225be00,0x225be08,0x225be10,0x225be00,0x225be10,0x225be10,0x225be00,0x225bd80,0x225bd80,0x224cc40,0x224cc41,0x224cc40,0x224cc41,0x224cc41,
0.02,0x225be00,0x225be08,0x225be10,0x225be00,0x225be10,0x225be10,0x225be00,0x225bd80,0x225bd80,0x224cc40,0x224cc41,0x224cc40,0x224cc41,0x224cc41,
0.024,0x225be00,0x225be08,0x225be10,0x225be00,0x225be10,0x225be10,0x225be00,0x225bd80,0x225bd80,0x224cc40,0x224cc41,0x224cc40,0x224cc41,0x224cc41,
0.028,0x225be00,0x225be08,0x225be10,0x225be00,0x225be10,0x225be10,0x225be00,0x225bd80,0x225bd80,0x224cc40,0x224cc41,0x224cc40,0x224cc41,0x224cc41,
0.032,0x225be00,0x225be08,0x225be10,0x225be00,0x225be10,0x225be10,0x225be00,0x225bd80,0x225bd80,0x224cc40,0x224cc41,0x224cc40,0x224cc41,0x224cc41,
0.036,0x225be00,0x225be08,0x225be10,0x225be00,0x225be10,0x225be10,0x225be00,0x225bd80,0x225bd80,0x224cc40,0x224cc41,0x224cc40,0x224cc41,0x224cc41,
0.04,0x225be00,0x225be08,0x225be10,0x225be00,0x225be10,0x225be10,0x225be00,0x225bd80,0x225bd80,0x224cc40,0x224cc41,0x224cc40,0x224cc41,0x224cc41,
0.044,0x225be00,0x225be08,0x225be10,0x225be00,0x225be10,0x225be10,0x225be00,0x225bd80,0x225bd80,0x224cc40,0x224cc41,0x224cc40,0x224cc41,0x224cc41,

[Niklas Worschech]

The csv output is not working at the moment. But the mat file output should work

[Martin Sjölund]

The nightly tests use CSV (I think because cppruntime did not support MAT).

[Niklas Worschech]

The cpp runtime supports mat file output. But I will also fix the csv output

[Niklas Worschech]

Is it possible to switch the test to mat file otuput, because the mat file output is a bit faster.

[Martin Sjölund]

I have tried to do so; started a new job. (The only reason the scripts can change the output format was to support c++runtime anyway; mat is default).

[Rüdiger Franke]

Many validations work again with mat files -- some still fail: for instance Modelica.Blocks.Examples.BooleanNetwork1. or2.y should be 0, but is -1!?

Is this an error in the mat file generation?

[Martin Sjölund]

I guess so. or2.y sounds like something that should be a boolean, so 0 or 1.

[Rüdiger Franke]

The errors occur for negated boolean aliases. Bool uses operator! instead of operator- for negation. This can become quite tricky in conjunction with implicit casts between bool and numeric types -- templates make live not only simpler :-(

f5433093d4a1dbd95a9f69e6d2cb75c6612e4577/OMCompiler introduces a specific WriteOutputVarBool.
3746b035ae0dfca69323cb7d6f216ffd3fab515f/OpenModelica-testsuite adds or2.y to the respective test.

[Niklas Worschech]

I will fix this, i thought negated boolean alias will never occur.

[Niklas Worschech]

Ahh you already fixed it :D. I din't not read that.
You can also use template specialization.

[Rüdiger Franke]

We are not yet back -- today 188 models verified. It were 194 before the breakdown!

See e.g.:

Modelica.Electrical.Spice3.Examples.CascodeCircuit
Modelica.Mechanics.MultiBody.Examples.Constraints.SphericalConstraint

[Martin Sjölund]

Milestone pushed to 1.9.5

[Martin Sjölund]

Milestone renamed

[Martin Sjölund]

Ticket retargeted after milestone closed

[Martin Sjölund]

Milestone moved to 1.12.0 due to 1.11.0 already being released.

[Francesco Casella]

The milestone of this ticket has been reassigned to "Future".

If you think the issue is still valid and relevant for you, please select milestone 1.13.0 for back-end, code generation and run-time issues, or 2.0.0 for front-end issues.

If you are aware that the problem is no longer present, please select the milestone corresponding to the version of OMC you used to check that, and set the status to "worksforme".

In both cases, a short informative comment would be welcome.

[Martin Sjölund]

This is fixed now. And we have a completely new testing environment anyway.