[llvm] [AArch64][SVE] Select immediate offsets for multi-vector instructions (PR #201637)

[Paul Walker via llvm-commits]

================
@@ -7715,6 +7677,14 @@ static EVT getPackedVectorTypeFromPredicateType(LLVMContext &Ctx, EVT PredVT,
   return MemVT;
 }
 
+/// Builds an integer vector type large enough to hold \p NumVec instances
+/// of \p VecVT.
+static EVT getPackedMultipleVectorType(LLVMContext &Ctx, EVT VecVT,
+                                       unsigned NumVec) {
+  return EVT::getVectorVT(Ctx, VecVT.getScalarType().changeTypeToInteger(),
----------------
paulwalker-arm wrote:

Or remove Packed from the name and keep it as a generic utility function.

https://github.com/llvm/llvm-project/pull/201637