zygoloid wrote: Instead of having two different kinds of diagnostic, we should mark the `ext_typecheck_zero_array_size` diagnostic as `SFINAEFailure`.
Also, why does this change make any difference? We should always get a SFINAE failure if we produce an `Error` diagnostic regardless of whether this function returns `QualType`. Is some enclosing context suppressing the diagnostic? The diagnostic we were getting is missing a backtrace, so I think there's actually a different bug here and this change is just hiding a symptom of it. https://github.com/llvm/llvm-project/pull/170144 _______________________________________________ cfe-commits mailing list [email protected] https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
