MuellerMP wrote: Imo this is a good fix in the sense that it reduces the emitted `scope.begin` + `scope.end`'s which are not handled by the state numbering algo. There are still some of those left in the IR code of the sample - would it make sense to investigate the use of those aswell? Regarding correctness this at least does not seem to cause regressions with windows_seh_tests (the microsoft github repo).
https://github.com/llvm/llvm-project/pull/180144 _______________________________________________ cfe-commits mailing list [email protected] https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
