alexfh added a comment. In https://reviews.llvm.org/D28729#647250, @Prazek wrote:
> Does solution like this works for you? We don't officially support alpha > checkers, but it is much easier to check if something is already implemented > in static analyzer easily Is it the only problem you're trying to solve by this patch? If so, it might be better to add an alias/script to run `clang --analyze -Xclang -analyzer-checker=alpha` or even use gcc.godbolt.org (e.g. https://godbolt.org/g/msQigz). My concern is that if there's a user-visible option (undocumented, unlisted, named `--no-no-no-please-dont-use-this-option` or otherwise discouraged), people will use it (sometimes unintentionally, because, well, they changed their config once to give it a try and then forgot about it for a year), rely on it, complain about the tool, etc. If clang-tidy is used in some sort of a server-side manner (e.g. integrated with code review), misbehaving checkers may cause crashes, and this is problematic, when maintainers have no intention to fix the bugs and well, they explicitly asked not to expose experimental checkers (see http://llvm.org/PR26855, for example). https://reviews.llvm.org/D28729 _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
