// Usually this happens automatically in TF_SessionRun. After this is called,
// TF_SessionRun will no longer extend the session on every call.
//
// We expose this here to allow fine-grained synchronization in multi-threaded
// workloads, which is required since the Python implementation depends on the
// above mutation methods. This allows us to prevent modifications to nodes in

///   1. A new method is added _at the end_ of the function table.
///   2. Preconditions or postconditions for one operation in these function
///   table change. Note that only core TensorFlow is able to impose these
///   invariants (i.e., guarantee the preconditions before calling the operation
///   and check the postconditions after the operation returns). If plugins need