[llvm] Fix scalar overload name constructed by ReplaceWithVeclib.cpp (PR #111095)

[Maciej Gabka via llvm-commits]

================
@@ -100,20 +100,39 @@ static void replaceWithTLIFunction(IntrinsicInst *II, VFInfo &Info,
 static bool replaceWithCallToVeclib(const TargetLibraryInfo &TLI,
                                     IntrinsicInst *II) {
   assert(II != nullptr && "Intrinsic cannot be null");
-  // At the moment VFABI assumes the return type is always widened unless it is
-  // a void type.
-  auto *VTy = dyn_cast<VectorType>(II->getType());
-  ElementCount EC(VTy ? VTy->getElementCount() : ElementCount::getFixed(0));
+  Intrinsic::ID IID = II->getIntrinsicID();
+  if (IID == Intrinsic::not_intrinsic)
+    return false;
+
+  // RetIsScalar: Return type is not widened.
+  bool RetIsScalar = isVectorIntrinsicWithScalarOpAtArg(IID, -1);
+  Type *RetTy = II->getType();
+  Type *ScalarRetTy = RetTy->getScalarType();
+
   // Compute the argument types of the corresponding scalar call and check that
   // all vector operands match the previously found EC.
   SmallVector<Type *, 8> ScalarArgTypes;
-  Intrinsic::ID IID = II->getIntrinsicID();
+  auto *VTy = dyn_cast<VectorType>(RetTy);
+  ElementCount EC(!RetIsScalar && VTy ? VTy->getElementCount()
----------------
mgabka wrote:

the extra !RetIsScalar is not needed, if the return type is vector then isVectorIntrinsicWithScalarOpAtArg(IID, -1) return false, the original place where EC was computed was correct, could you please revert this change?

https://github.com/llvm/llvm-project/pull/111095