kadircet wrote: Hi folks, we're seeing some regressions on our internal tidy checks after this patch.
The regression is similar to what @haoNoQ raised in https://github.com/llvm/llvm-project/pull/128150#issuecomment-2705425193, the blast radius a little bit wider than checks that build and internal database though. `TraversalScope` also controls how widely used matchers like `hasAncestor/hasParent` behaves (or any other direct uses of a `RecursiveASTVisitor`). Hence any tidy check that looks for parents of a declaration via those helpers will actually be affected by this change and will require adjustments similar to https://github.com/llvm/llvm-project/pull/128150/files#diff-2fbcf6302e9123918381fc45c1bc4a9dc9208e4a1d2166c5315882fa93122ac5. I wanted to make sure we're still OK with this change (looking at the thread, that sounds like the case, but I wanted to emphasize that the blast radius is much wider than checks analyzing declarations inside system headers. it's also affecting checks that analyzes declarations that refer to system headers). ATM, this behavior is coupled with emitting findings for system headers. Can we at least introduce a new option to control this new behavior? https://github.com/llvm/llvm-project/pull/128150 _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
