[AArch64] Tweak the costs of experimental_cttz_elts intrinsic (#125093)

The experimental_cttz_elts intrinsic currently returns a cost
of 1 for all types, however we know that it currently requires
2 SVE instructions when lowering this - brkb and cntp. Both of
these instructions have a throughput that is half of a basic
vector instruction such as a vector add. This patch bumps the
cost of this intrinsic up to 4 to reflect two instructions of
lower throughput.

Author: david-arm

llvm/lib/Target/AArch64/AArch64TargetTransformInfo.cpp

@@ -940,6 +940,16 @@ AArch64TTIImpl::getIntrinsicInstrCost(const IntrinsicCostAttributes &ICA,
     }
     break;
   }
+  case Intrinsic::experimental_cttz_elts: {
+    EVT ArgVT = getTLI()->getValueType(DL, ICA.getArgTypes()[0]);
+    if (!getTLI()->shouldExpandCttzElements(ArgVT)) {
+      // This will consist of a SVE brkb and a cntp instruction. These
+      // typically have the same latency and half the throughput as a vector
+      // add instruction.
+      return 4;
+    }
+    break;
+  }
   default:
     break;
   }