[new release] bitwuzla-cxx (0.6.0)

[recoules]

SMT solver for AUFBVFP (C++ API)

• Project page: https://bitwuzla.github.io
• Documentation: https://bitwuzla.github.io/docs/ocaml/

CHANGES:

Update Bitwuzla sources.

Fix a segmentation fault that occurs during the garbage collection of node managers.

Vendor submodules:

• Bitwuzla tag:0.6.0

[shonfeder]

Hi! Thanks for publishing this update.

In CI we have the following failures:

arm32-ocaml-4.14
bitwuzla-cxx.0.6.0
tests (failed: The compilation of bitwuzla-cxx.0.6.0 failed at "dune build -p bitwuzla-cxx -j 79 @install @runtest".)
arm32-ocaml-5.2
bitwuzla-cxx.0.6.0
tests (failed: The compilation of bitwuzla-cxx.0.6.0 failed at "dune build -p bitwuzla-cxx -j 79 @install @runtest".)

Do you intend to support these platforms?

[recoules]

Hi @shonfeder,

I do not see why there is a bus error on arm32.
I do not know either if previous versions were tested for this architecture too (regression) or if it never worked on it.

Yet, I do not have easy access to this kind of architecture and I will not have time to investigate (for instance, if the issue is related to the bitwuzla original c++ code or only to the binding). Thus, I propose to add an opam filter for arm32.

Best regards,
Frédéric Recoules

PS: seems the previous version had the same issue, and the same filter.

[shonfeder]

PS: seems the previous version had the same issue, and the same filter.

I see! Perhaps this should be upstreamed to the package data in the source?

Thanks for the fix here! Despite the github UI's report, all CI jobs appear to be passing!

[shonfeder]

Thanks again for taking the time to share your updates! 🙏

Please consider announcing on changes on discuss.ocaml.org :)

[recoules]

Thank you.

Perhaps this should be upstreamed to the package data in the source?

I updated the opam.template file in the source repository.

[shonfeder]

Excellent. Thanks!