ParentFull threadShoop·Is there a way to generate IR which is easier on LLVM? The Jai compiler compiles 60kloc with an llvm backend in seconds. Could performance like this be seen in rustc?View on HN