Szelethus added a comment. Hmmm, I was wrong a little bit: I realized that if I tinker with with `initializeManager` and `getEnabledCheckers` just a bit more, register checkers straight away, don't need to collect them first, and this would make sure that dependencies are registered before the checkers that depend on them. That means I can get away this patch without even touching the MallocChecker family. `CheckerManager` would receive the `getChecker` function that would assert if the checker isn't already registered, so we could be extra sure this system works.
Buuuut since I spent a lot of time with MallocChecker, I might tickle the splitting problem later on as a weekend project or something. Repository: rC Clang CHANGES SINCE LAST ACTION https://reviews.llvm.org/D54438/new/ https://reviews.llvm.org/D54438 _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
