Reply To: The Belt

> However, argument reordering is not done by LLVM, and we have no plans to add it to LLVM any time soon.

But doesn’t conform loop quite a bit like the block arguments that MLIR uses, as a different way to represent PHI nodes? It seems that conform is a sensible way to model branches that have multiple target PHI nodes. (I have programmed LLVM with such PHI nodes, but I understand why MLIR uses argument instead).