I’m not sure if this is the place to ask, but after I add ghc-bignum to my library’s depends. It fails to build under GHC 8.X. cabal complains ghc-bignum is not buildable.
So the question is I’m not sure about the migration path for code using GMP internals, should I use a CPP to import different constructors from different versions of integer-gmp?
ghc-bignum isn’t supposed to be used by GHC < 9.0. GHC 9.0 should come with integer-gmp 1.1 that provides backward compatible pattern synonyms. Otherwise if you really need to import something from ghc-bignum, you will probably have to use something like CPP in code and if impl(ghc > 9.0) in .cabal files.