danix800 added a comment. In D156461#4583806 <https://reviews.llvm.org/D156461#4583806>, @balazske wrote:
> Normally what should happen is that a new move constructor is imported (with > a definition) and linked after the existing one (and the existing is not > modified). Does this violate the constraint that ctor shouldn't be re-declared? > We get an AST that does not occur after a normal compile, I do not know if > this causes problems or if this is the real reason for this patch. Agreed. The original AST should not be touched. > What should be done is find the existing constructor and update it with the > definition and return it from the import. This can be done with any type of > constructor. So does this mean that importer might be improved on this part? Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D156461/new/ https://reviews.llvm.org/D156461 _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
