[llvm] [AArch64] Improve cost model for legal subvec insert/extract (PR #81135)

[Sander de Smalen via llvm-commits]

================
@@ -568,6 +568,48 @@ AArch64TTIImpl::getIntrinsicInstrCost(const IntrinsicCostAttributes &ICA,
     }
     return Cost;
   }
+  case Intrinsic::vector_extract: {
+    // If both the vector argument and the return type are legal types and the
+    // index is 0, then this should be a no-op or simple operation; return a
+    // relatively low cost.
+
+    // If arguments aren't actually supplied, then we cannot determine the
+    // value of the index.
+    if (ICA.getArgs().size() < 2)
+      break;
+    LLVMContext &C = RetTy->getContext();
+    EVT MRTy = getTLI()->getValueType(DL, RetTy);
+    EVT MPTy = getTLI()->getValueType(DL, ICA.getArgTypes()[0]);
+    TargetLoweringBase::LegalizeKind RLK = getTLI()->getTypeConversion(C, MRTy);
+    TargetLoweringBase::LegalizeKind PLK = getTLI()->getTypeConversion(C, MPTy);
+    const ConstantInt *Idx = dyn_cast<ConstantInt>(ICA.getArgs()[1]);
+    if (RLK.first == TargetLoweringBase::TypeLegal &&
+        PLK.first == TargetLoweringBase::TypeLegal && Idx &&
+        Idx->getZExtValue() == 0)
+      return InstructionCost(1);
----------------
sdesmalen-arm wrote:

I guess the extra instructions required for extracting from an unpacked vector is `2log(SVEBitsPerBlock/(MinNumElts * EltSizeInBits))`. For the insert that would require adding another `TCC_Basic` for the cost of the mov (when it's not inserting into an undef/poison value). This would automatically return `TCC_Free` (==`0`) if the scalable vector is a packed vector.

https://github.com/llvm/llvm-project/pull/81135