arsenm wrote: > This is not a "real" regression though, right? As in, we'd already get that > behavior for the IR clang actually generates (with range attributes on the > call-sites): https://llvm.godbolt.org/z/GWcs9837a It's just that the codegen > tests weren't actually specifying the attribute that clang produces?
Yes, for the clang case. I don't think other frontends have been bothering to set this https://github.com/llvm/llvm-project/pull/136196 _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
