[llvm] [AArch64][ISel] Prefer SVE over NEON for CTLZ (PR #194839)

[Benjamin Maxwell via llvm-commits]

================
@@ -1931,9 +1931,11 @@ AArch64TargetLowering::AArch64TargetLowering(const TargetMachine &TM,
       setOperationAction(ISD::MULHU, VT, Custom);
     }
 
-    // NEON doesn't support 64-bit vector integer muls, but SVE does.
+    // NEON doesn't support 64-bit vector integers, but SVE does.
----------------
MacDue wrote:

nit:
```suggestion
    // NEON doesn't support 64-bit vector integer mul and ctlz operations, but SVE does.
```

https://github.com/llvm/llvm-project/pull/194839