0x8000-0000 added a comment. In D54943#1815568 <https://reviews.llvm.org/D54943#1815568>, @JonasToth wrote:
> @0x8000-0000 i did remove `dependentTypes` from matching. Locally that works > with your reported test-case (i was running clang-tidy wrong :(). > > I do understand the `auto &` issues now (i think). > If the type is deduced to a template-type or something that depends the > standard ways of detecting that do not work, because in each template > instantiation this information is not available. > > Having a `auto & local = TemplatedVariable` does not have the > `isInstantiationDependentType()`-bits set and for each instantiation of the > template the const-analysis is run. > The false positives happened in different template instantiations, i think > because the overloaded operators were sometimes `const` and sometimes not. I > tried many ways to detect, if the type comes from a template, but had no > success. > This is probably an issue in `Sema` or requires adjustments there. > I added tests in clang-tidy, but `#if 0` them out because are not working > right now. Thank you for looking into this; I'll start building diff14 to test locally. I'm good with merging this checker as-is, as long as it is not enabled by default. I want to be able to use it even partially, in environments where I might not be able to inject a home-built compiler ;) Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D54943/new/ https://reviews.llvm.org/D54943 _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
