[llvm] [AArch64] Optimize vector multiplications by certain constants for v2i64 (PR #183827)

[David Green via llvm-commits]

================
@@ -5965,7 +6003,8 @@ SDValue AArch64TargetLowering::LowerMUL(SDValue Op, SelectionDAG &DAG) const {
       // legal.
       if (Subtarget->hasSVE())
         return LowerToPredicatedOp(Op, DAG, AArch64ISD::MUL_PRED);
-      // Fall through to expand this.  It is not legal.
----------------
davemgreen wrote:

Leave a comment that explains that other v2i64 muls are not legal so expanded.

https://github.com/llvm/llvm-project/pull/183827