martinboehme wrote: Update: I've run some tests on an internal codebase, and with this PR, Crubit's nullability check sees a reduction in SAT solver timeouts of over 40% and a reduction in "reached maximum iterations" errors of over 60% -- similar to the previous more complicated approach in #72850.
https://github.com/llvm/llvm-project/pull/72985 _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
