[llvm] [AArch64] Fix SVE cost model for various math intrinsics (PR #184358)

[David Sherwood via llvm-commits]

================
@@ -696,9 +696,10 @@ AArch64TTIImpl::getIntrinsicInstrCost(const IntrinsicCostAttributes &ICA,
     break;
   }
   case Intrinsic::abs: {
-    static const auto ValidAbsTys = {MVT::v8i8,  MVT::v16i8, MVT::v4i16,
-                                     MVT::v8i16, MVT::v2i32, MVT::v4i32,
-                                     MVT::v2i64};
+    static const auto ValidAbsTys = {MVT::v8i8,    MVT::v16i8,   MVT::v4i16,
+                                     MVT::v8i16,   MVT::v2i32,   MVT::v4i32,
+                                     MVT::v2i64,   MVT::nxv16i8, MVT::nxv8i16,
+                                     MVT::nxv4i32, MVT::nxv2i64};
----------------
david-arm wrote:

For consistency with the NEON behaviour it's easier if I just add the scalable vector types to this table. In the absence of SVE either the type won't match or we'll just return Invalid cost, which I confirmed with a downstream test.

https://github.com/llvm/llvm-project/pull/184358