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

[David Green via llvm-commits]

================
@@ -1114,6 +1121,25 @@ AArch64TTIImpl::getIntrinsicInstrCost(const IntrinsicCostAttributes &ICA,
       return getCallInstrCost(nullptr, RetTy, ICA.getArgTypes(), CostKind);
     break;
   }
+  case Intrinsic::sqrt:
----------------
davemgreen wrote:

Vector sqrt shouldn't have a throughput of 1, IIUC, but that might be true of neon too and div - we don't model them very accurately at the moment.

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