[llvm] [AArch64] Scalarize v2f16 vecreduce.fadd (PR #147783)

[Sam Tebbs via llvm-commits]

================
@@ -16046,9 +16049,19 @@ static SDValue getVectorBitwiseReduce(unsigned Opcode, SDValue Vec, EVT VT,
 SDValue AArch64TargetLowering::LowerVECREDUCE(SDValue Op,
                                               SelectionDAG &DAG) const {
   SDValue Src = Op.getOperand(0);
+  EVT SrcVT = Src.getValueType();
+
+  // Scalarize v2f32 to turn it into a faddp. This will be more efficient than
----------------
SamTebbs33 wrote:

v2f32 -> v2f16 ?

https://github.com/llvm/llvm-project/pull/147783