erichkeane marked 2 inline comments as done. erichkeane added inline comments.
================ Comment at: clang/include/clang-c/Index.h:202-204 + * The cursor has a declspec(nothrow) exception specification. + */ + CXCursor_ExceptionSpecificationKind_NoThrow, ---------------- rsmith wrote: > This would renumber the later enumerators, resulting in an ABI break for our > stable C ABI. Is it alright to just add it to the end then? Or do I need to translate it into one of the others? ================ Comment at: clang/lib/Sema/SemaExprCXX.cpp:6067-6068 if (EST2 == EST_NoexceptTrue) return ESI1; + if (EST1 == EST_NoThrow) return ESI2; + if (EST2 == EST_NoThrow) return ESI1; ---------------- rsmith wrote: > I think this should be checked earlier, in the spirit of allowing "real" > syntax to be preferred to attributes. (Eg, given a declaration that's > `__declspec(nothrow)` and one that's `noexcept`, we should keep the > `noexcept` in the merged type.) Ah, right, good point. Looking above, I'm torn between above or below MSAny. Do you have a thought? Repository: rC Clang CHANGES SINCE LAST ACTION https://reviews.llvm.org/D62435/new/ https://reviews.llvm.org/D62435 _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits