5chmidti wrote: > I think the checker can be removed if clang has this kind of compilation time > warning and can totally cover the checker.
It only covers the whole check if `StrictMode` is not set to `false`, because that enables diagnosing down-casts for non-polymorphic types as well. > Perhaps that check should ensure this warning is enabled and otherwise do > nothing? Afaik that would be a first. If there is an equivalent warning for a check, then the check can be removed I favor of the warning. In this case, I think that the check is probably going to stay, but I think that will be a question for the time when this PR is finalized/merged, as to how much the warning covers the check. @cor3ntin with CRTP there shouldn't be virtual functions, otherwise why use CRTP, which means the record fails the check for being polymorphic. https://github.com/llvm/llvm-project/pull/117914 _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
