[llvm] [LV] Vectorize FMax via OrderedFCmpSelect w/o fast-math flags. (PR #146711)

[via llvm-commits]

================
@@ -937,6 +937,7 @@ constexpr Intrinsic::ID llvm::getReductionIntrinsicID(RecurKind RK) {
     return Intrinsic::vector_reduce_umax;
   case RecurKind::UMin:
     return Intrinsic::vector_reduce_umin;
+  case RecurKind::OrderedFCmpSelect:
----------------
ayalz wrote:

This is for max, returning Intrinsic::vector_reduce_fmax; need a separate case for min, returning Intrinsic::vector_reduce_fmin as below?

https://github.com/llvm/llvm-project/pull/146711