tbaeder added a comment. In D141472#4217479 <https://reviews.llvm.org/D141472#4217479>, @aaron.ballman wrote:
> In D141472#4112790 <https://reviews.llvm.org/D141472#4112790>, @tbaeder wrote: > >> I'm not sure what to do about this right now. I was wondering about >> restructuring pointers so all the metadata is before all the actual data, >> but that would be a large refactoring. > > For right now, we can move forward with basic function pointer support and we > can revisit this when you get around to doing member function pointers. FWIW, I did that refactoring locally but it didn't really turn out to work like that. This example works now and there should be a test for it AFAIK. ================ Comment at: clang/lib/AST/Interp/ByteCodeExprGen.cpp:970 return this->emitNullPtr(E); + case PT_FnPtr: case PT_Float: ---------------- aaron.ballman wrote: > Shouldn't this call `emitNullPtr()` in this case? e.g., > ``` > struct S { > int i; > void (*fp)(); > }; > > constexpr S s{ 12 }; > static_assert(s.fp == nullptr); > ``` Heh, that function is dead code right now, I keep it around because I think I might need it later. But your reproducer is broken anyway, I'll fix it, thanks. CHANGES SINCE LAST ACTION https://reviews.llvm.org/D141472/new/ https://reviews.llvm.org/D141472 _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
