Record binding missing causing string to be assigned 0.0

[ali.shirazi@…]

Hi there,

After upgrading to the new version of omc and omlibraries, we found a compiling-related error while running a test case () from our open-modelica based repositort, SOLARTHERM, which can be found from ​https://github.com/SolarTherm/SolarTherm. The instructions to run this test case is given in ​https://github.com/SolarTherm/SolarTherm/wiki/Building-SolarTherm.

I am copying the error in the following:

[/build/openmodelica-djNULe/openmodelica-1.12.0/OMCompiler/Compiler/BackEnd/Unit.mo:869:7-869:68:writable] Error: Internal error function lexer failed

Compiling simulator
SimpleSystem.c:1135:91: warning: magnitude of floating-point constant too large
      for type 'double'; maximum is 1.7976931348623157E+308 [-Wliteral-range]
  .../* pri._nextTimeEventScaled DISCRETE */,1.797693134862316e+308);
                                             ^
SimpleSystem.c:1136:154: warning: magnitude of floating-point constant too large
      for type 'double'; maximum is 1.7976931348623157E+308 [-Wliteral-range]
  .../* pri._nextTimeEventScaled DISCRETE */:1.797693134862316e+308);
                                             ^
2 warnings generated.
SimpleSystem_06inz.c:468:88: warning: magnitude of floating-point constant too
      large for type 'double'; maximum is 1.7976931348623157E+308
      [-Wliteral-range]
  .../* pri._nextTimeEventScaled DISCRETE */,1.797693134862316e+308);
                                             ^
SimpleSystem_06inz.c:469:151: warning: magnitude of floating-point constant too
      large for type 'double'; maximum is 1.7976931348623157E+308
      [-Wliteral-range]
  .../* pri._nextTimeEventScaled DISCRETE */:1.797693134862316e+308);
                                             ^
2 warnings generated.
SimpleSystem_08bnd.c:675:44: error: assigning to 'modelica_string'
      (aka 'void *') from incompatible type 'double'
  data->simulationInfo->stringParameter[6] = 0.0;
                                           ^ ~~~
1 error generated.
<builtin>: recipe for target 'SimpleSystem_08bnd.o' failed
make: *** [SimpleSystem_08bnd.o] Error 1
Traceback (most recent call last):
  File "/home/ali/.local/bin/st_simulate", line 103, in <module>
    sim.compile_sim(args=([] if args.v else ['-s']))
  File "/home/ali/.local/lib/python2.7/site-packages/solartherm/simulation.py", line 244, in compile_sim
    + args)
  File "/usr/lib/python2.7/subprocess.py", line 541, in check_call
    raise CalledProcessError(retcode, cmd)
subprocess.CalledProcessError: Command '['make', '-j', '1', '-f', 'SimpleSystem.makefile', '-s']' returned non-zero exit status 2

Apparently, the issue is related to the Strings.substring command which is supposed to find a string from a file, but it sets a float value instead, and this raises an error. I should also mention that the older compiler versions were working just fine. As such we suspect that there is some bug in the compiling process in the omc v1.12.

I would greatly appreciate it if you could have a look at this issue and let us know what is exactly wrong.

Regards,

Ali Shirazi

[ali.shirazi@…]

Further to this issue, I need to clarify that our test case has been run through terminal in Ubuntu 16.04.3 64 bit.

[Martin Sjölund]

Looks like a frontend issue. In 1.11, the record was evaluated at compile-time and replaced with its values; in 1.12 we changed some things regarding records and it seems probably the evaluation fails. In the master, wea.meta = SolarTherm.Sources.Weather.getMetadata(wea.file_act); appears as an initial equation which should have helped:

[/home/marsj/solartherm/SolarTherm/Sources/Weather.mo:21:2-21:49:writable] Notification: Moving binding to initial equation section and setting fixed attribute of wea.meta.name to false.
[/home/marsj/solartherm/SolarTherm/Sources/Weather.mo:21:2-21:49:writable] Notification: Moving binding to initial equation section and setting fixed attribute of wea.meta.lat to false.
[/home/marsj/solartherm/SolarTherm/Sources/Weather.mo:21:2-21:49:writable] Notification: Moving binding to initial equation section and setting fixed attribute of wea.meta.lon to false.
[/home/marsj/solartherm/SolarTherm/Sources/Weather.mo:21:2-21:49:writable] Notification: Moving binding to initial equation section and setting fixed attribute of wea.meta.elev to false.
[/home/marsj/solartherm/SolarTherm/Sources/Weather.mo:21:2-21:49:writable] Notification: Moving binding to initial equation section and setting fixed attribute of wea.meta.tzone to false.
[/home/marsj/solartherm/SolarTherm/Sources/Weather.mo:21:2-21:49:writable] Notification: Moving binding to initial equation section and setting fixed attribute of wea.meta.tstart to false.

But we also get:

[/home/marsj/solartherm/SolarTherm/Sources/Weather.mo:92:2-92:28:writable] Warning: Parameter wea.meta.name has no value, and is fixed during initialization (fixed=true), using available start value (start=0.0) as default value.
Error: The given system is mixed-determined.   [index > 3]
Please checkout the option "--maxMixedDeterminedIndex".

So I need to see what's going on there. It's unlikely to be fixed in a 1.12 maintenance release.

[Martin Sjölund]

Fixed by #4703. It would require some larger changes to the 1.12 release so I suggest using 1.11 or 1.13 nightly builds.