Remove caching of return slot usage. (#4163)

The caching isn't buying us much, and is adding complexity and
divergence between the codepaths for generic and non-generic functions.

This means we no longer suppress diagnostics for the second or
subsequent time we call a function with an incomplete return type. If we
want to add that back, it might be worth considering moving the
suppression to `TryToCompleteType` and only diagnosing that a type is
incomplete once, regardless of why we're requiring it to be complete.

toolchain/sem_ir/function.cpp

@@ -51,4 +51,20 @@ auto Function::GetDeclaredReturnType(const File& file,
                            file.insts().Get(return_storage_id).type_id());
 }
 
+auto ReturnInfo::ForType(const File& file, TypeId type_id) -> ReturnInfo {
+  if (!type_id.is_valid()) {
+    // Implicit `-> ()` has no return slot.
+    return {.type_id = type_id, .return_slot = ReturnSlot::Absent};
+  }
+
+  if (!file.types().IsComplete(type_id)) {
+    return {.type_id = type_id, .return_slot = ReturnSlot::Incomplete};
+  }
+
+  return {.type_id = type_id,
+          .return_slot = GetInitRepr(file, type_id).has_return_slot()
+                             ? SemIR::ReturnSlot::Present
+                             : SemIR::ReturnSlot::Absent};
+}
+
 }  // namespace Carbon::SemIR