erichkeane wrote: > Hi all, just checking in on the preferred direction here. Should I continue > with the defensive diagnostic approach in this PR, or would you prefer that I > investigate a compact zero-filled representation for sparse designated > initializers instead?
Ideally we'd do figure out a way to take advantage of everything LLVM lets us do, but frankly, I suspect that LLVM is already going to be a quite limiting factor with quite a lot of work here. I think the idea of a limit on these lists is reasonable, but as above folks have said, we need this to be command-line-configurable. we have similar imp-limit config options. https://github.com/llvm/llvm-project/pull/205503 _______________________________________________ cfe-commits mailing list [email protected] https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
