alexfh added a comment. In https://reviews.llvm.org/D25659#589081, @malcolm.parsons wrote:
> In https://reviews.llvm.org/D25659#588663, @alexfh wrote: > > > In https://reviews.llvm.org/D25659#588658, @alexfh wrote: > > > > > I think, silently choosing one of the checks may be confusing and > > > counter-intuitive. Should we just warn in case we see the same check > > > enabled by multiple aliases? > > > > > > And by "aliases" I mean different names under which the check is > > registered. I'm not sure I see the benefit of introducing a separate > > concept and a separate API for registering checks under multiple names. > > This introduces another sort of dynamic binding, which we have to have good > > reasons for. > > > Is using `typeid()` on the constructed checks OK? I don't think rtti is allowed in LLVM. But we can find a solution, I believe. The important thing is to decide what behavior is desired. https://reviews.llvm.org/D25659 _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits