ABI of the library depends on the value of NDEBUG?

[traversaro]

It seems that currently the ABI of the library depends on the value of NDEBUG used when using the headers of the library, see:

• dart/dart/constraint/PgsBoxedLcpSolver.hpp

  Line 83 in 2437749

  bool canSolve(int n, const double* A) override;

• dart/dart/constraint/PGSLCPSolver.hpp

  Line 61 in 2437749

  #ifndef NDEBUG

• dart/dart/common/MemoryManager.hpp

  Line 154 in 2437749

  #ifndef NDEBUG

However, this value could be different in the downstream project using dart w.r.t. to the value used when compiling dart, for example see conda-forge/gz-physics-feedstock#26 (comment) .

Probably it could make sense to save in

dart/dart/config.hpp.in

Line 65 in 2437749

#cmakedefine01 BUILD_TYPE_DEBUG

the NDEBUG value used when compiling dart (or something similar) and use that in the .hpp ?

Bug Report

Please answer the following questions for yourself before reporting a bug.

• I checked the documentation and the forum but found no answer.
• I checked to make sure that this issue has not already been filed.

Environment

Select the following information.

• DART version: 6.13.2
• OS name and version name(or number): Not relevant
• Compiler name and version number: Not relevant

Expected Behavior

Please describe the behavior you are expecting.

Current Behavior

What is the current behavior?

See description.

Steps to Reproduce

Not relevant?

Code to Reproduce

Please remember that with a working code it's easier to reproduce the bug and it's much faster to fix it.
It would be great to fork this repo and create a regression test using this template.

Not relevant?

[jslee02]

Good suggestion! Created #1796

[jslee02]

I believe #1796 has resolved this issue. If there are more things that need to be resolved, please reopen it or create a new one. Thank you!