jhuber6 wrote: > I imagine the better fix would be to either overload the intrinsic or make it > return a pointer in the global address space, depending on whether it's > semantics tied to globals or this is just an implementation choice.
It's definitely tied to globals for `noop` in this case, so I did what you suggested. I think we should apply this to a lot of the other ones, but they should be separate patches and this solves my immediate problem. https://github.com/llvm/llvm-project/pull/102096 _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
