https://github.com/usx95 commented:
> Something for future PRs: I think it would be nice to have some sort of > coverage statistics. While initially you do not aim to cover all of C++, it > would be good to know what portion of nodes are missing from the fact > generation when we run this on some real world code. That can give us some > progress indicators. Good idea. I was thinking along similar lines and have some ideas for initial metrics: - Use an RAV to visit all expressions in a function and count number of expressions which should have an origin (i.e., it is of pointer type) but does not have an origin associated with the expression. - Count origins that are created but are never associated with a Loan; this would be a signal that we don't understand the pointer's source. I will add these in a follow-up patch. https://github.com/llvm/llvm-project/pull/142313 _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
