[llvm] [LLVM][CodeGen] Add lowering for scalable vector bfloat operations. (PR #109803)

[Paul Walker via llvm-commits]

================
@@ -1070,13 +1071,21 @@ void VectorLegalizer::Expand(SDNode *Node, SmallVectorImpl<SDValue> &Results) {
     break;
   case ISD::FMINNUM:
   case ISD::FMAXNUM:
+    if (SDValue Expanded = ExpandBF16Arith(Node)) {
----------------
paulwalker-arm wrote:

I've moved the expansion function into TargetLowering and updated the two expandFMIN###_FMAXNUM### functions to make us of it.

https://github.com/llvm/llvm-project/pull/109803