New FMU (OMEdit 1.13.2) failed load from cffi in lisp

[FBXon2ROX]

Thank you for your great work.

I am working on Debian 64bit on a personal project creating a lisp package that loads FMU using cffi interface (similar to the python). For info I am working with emacs+slime+sbcl last year.

This all worked well so far and managed to connect to the fmu instance and fmi2-get-real or get-differential :-) if I may say.

The instantiation of a model exchange interface worked fine until some modifications in the FMU export (I am working with me and cs exports) in OMEdit 1.13.2.

Thank you for your support

I am not asking about lisp and cffi here as it already worked with a previous fmu in an older OMEdit but could you give me some ideas about what has been modified in the new version ?

I have checked the fmi2Functions.h of both FMUs the old and the new and nothing has changed here. The fmu2_model_interface.c has changed.
Are you aware of any issues that might help me here to go a little bit further ?

PS : Sorry I didn't track the previous OMEdit version that used to work :-( but it was definitely working last year.

[anonymous]

First order is not working

[anonymous]

2nd order is working

[Adrian Pop]

What is the actual problem? What is the error? Not working how?

• Does the .so in the fmu is loaded properly in the process?
• Does the FMU instantiates fine?

[anonymous]

Replying to adrpo:

What is the actual problem? What is the error? Not working how?

• Does the .so in the fmu is loaded properly in the process?
• Does the FMU instantiates fine?

Sorry for the late reply but I was expecting an email warning .

Well it seems that warnings are somehow erratic or even inexistent. A late attempt lead to the following warnings ... with missing access to LAPACK functions.

In the following case the old fmu works fine.

Here are the first lines of the warning :
The first line is my command line to create a model exchange object from another instance containing the description.xml data. The following are the warning messages.

So my guess is that the new .so file doesn't point or contain the lapack functions (? not sure)
REMARK: By the way it's emacs+slime+sbcl

BEGINING OF THE TEST ERROR MESSAGE
FMI> (defparameter fmu-test (make-instance-me fmu-info ))

STYLE-WARNING: Undefined alien: "cgbrfsx_"
STYLE-WARNING: Undefined alien: "cgbsvxx_"
STYLE-WARNING: Undefined alien: "cgerfsx_"
STYLE-WARNING: Undefined alien: "cgesvxx_"
STYLE-WARNING: Undefined alien: "cherfsx_"
STYLE-WARNING: Undefined alien: "chesvxx_"
STYLE-WARNING: Undefined alien: "cla_gbamv_"
STYLE-WARNING: Undefined alien: "cla_gbrcond_c_"
STYLE-WARNING: Undefined alien: "cla_gbrcond_x_"
STYLE-WARNING: Undefined alien: "cla_gbrfsx_extended_"
STYLE-WARNING: Undefined alien: "cla_gbrpvgrw_"
STYLE-WARNING: Undefined alien: "cla_geamv_"
STYLE-WARNING: Undefined alien: "cla_gercond_c_"
STYLE-WARNING: Undefined alien: "cla_gercond_x_"
STYLE-WARNING: Undefined alien: "cla_gerfsx_extended_" ........

[FBXon2ROX]

FMU access working session

[Adrian Pop]

How did you compiled the FMU?
From OMEdit Tools->Options->FMI, choose Platform: Static.

[FBXon2ROX]

Detailed steps fo the callbackfunctions fmu instantiation (static not working case)

[FBXon2ROX]

Replying to adrpo:

How did you compiled the FMU?
From OMEdit Tools->Options->FMI, choose Platform: Static.

Seems like my previous answer didn't register.

I have used both type of compilation and labriry format static and dynamic.

Please also find for your information a detailed description of the callback functions fmu instantiation process.

[FBXon2ROX]

Another info for you ?

I can load the new fmu in OMEdit and use it in a model.
So it seems that any changes are followed by the OMEdit protocol and working.
I have no options to test the fmu loading process with another fmu loader.
Could you recommend a tool I may use in order to check that the fmu produced is correct ?
I gladly could use it to give you more data.

Thank you for you support

[FBXon2ROX]

The previous version that used to work is OMEdit v 1.12.0

[Martin Sjölund]

Do you get those lapack warnings for the attached FMUs? They do not have any references to these functions, so the message shouldn't appear. OMSimulator also loads the FMUs just fine.

From your txt-file it seems that the error is that the simulation just freezes? I think 1.12.0 already contained some of the fixes for making sure the garbage collector never runs, and to statically link the executable by default.

[Francesco Casella]

Releasing 1.14.0 which is stable and has many improvements w.r.t. 1.13.2. This issue is rescheduled to 1.16.0

[Francesco Casella]

Retargeted to 1.17.0 after 1.16.0 release

[Francesco Casella]

Retargeted to 1.18.0 because of 1.17.0 timed release.

[Francesco Casella]

Ticket retargeted after milestone closed