C2131 expression did not evaluate to a constraint

[darkpotpot]

Hi,

I'm currently updating my projects in order to use c++20 and I'm having trouble with one of them. It uses boost geometry (which uses boost multiprecision), and at some point in the compilation it failed with the following error:

C2131 expression did not evaluate to a constraint

If I dig a little bit, the error comes from 2 files :

• traits\max_digits10.hpp:
  • l74 : static constexpr std::size_t value = digits_10(digits);
  • l44: static constexpr std::size_t value = max_digits_10(digits);
• cpp_int\limits.hpp:
  • l188: static constexpr int digits10 = static_cast(boost::multiprecision::detail::calc_digits10_s<static_cast(digits)>::value);
  • l189: static constexpr int max_digits10 = static_cast(boost::multiprecision::detail::calc_max_digits10_s<static_cast(digits)>::value);

I'm using Visual studio 2022 with v142 "platform toolset" and ISO C++20 "language standard" Boost version is 1.83.0

Does anybody know how I can get rid of the issue ?

Regards

Julien

Edit : I wrongly suspected language standard. It fails with all. During the update I switched from boost 1.72 to 1.83 because of issues in c++20 with others boost modules. So in fact it came from this update, I tested several version and issue appears in 1.78.0. All release after this one shows the issue Version 1.77.0 and previous are running fine

[jzmaddock]

That's strange, we obviously test with MSVC (and I use for development here), and I haven't seen that. Do you have a reduced self-contained test case I can look at?

[ckormanyos]

using Visual studio 2022 with v142

This one thing struck me as a bit unconventional... 2022 ships with 143. I don't se why this could be responsible for strange behavior but it is a non-standard config?

[darkpotpot]

Thanks for the answer, I'll try reproducing it with a minimal example and post it here

[darkpotpot]

Hm, culprit found. It was caused by another 3rd party lib (APDFL). Sorry for the trouble