// scheduled on that same threadpool, causing a deadlock in cases where the
  // caller of event_mgr->ThenExecute() blocks on the completion of the callback
  // (as in the case of ConstOp kernel creation on GPU, which involves copying a
  // CPU tensor to GPU).
  // Setting a larger thread pool does not help with the Swift caller, as we use
  // a different TFE context for each thread of execution (for running graph

  // Fetches the current Jacobian-vector product associated with `tensor_id`, or
  // a nullptr if none is available.
  //
  // Returns a borrowed reference, i.e. does not run VSpace::MarkAsResult on its
  // return value. The caller should increment the reference count before
  // deleting the ForwardAccumulator or calling DeleteGradient if keeping a
  // persistent reference to a non-null result.
  Gradient* FetchJVP(int64_t tensor_id);

Aaron Schumacher, Aidan Dang, Akihiko ITOH, Aki Sukegawa, Arbit Chen, Aziz Alto,
Danijar Hafner, Erik Erwitt, Fabrizio Milo, Felix Maximilian Möller, Henry
Saputra, Sung Kim, Igor Babuschkin, Jan Zikes, Jeremy Barnes, Jesper Steen
Møller, Johannes Mayer, Justin Harris, Kashif Rasul, Kevin Robinson, Loo Rong
Jie, Lucas Moura, Łukasz Bieniasz-Krzywiec, Mario Cho, Maxim Grechkin, Michael

      "Creating while loops is not supported on mobile. File a bug at "
      "https://github.com/tensorflow/tensorflow/issues if this feature is "
      "important to you");
#else
  // If it appears the caller created or modified `params`, don't free resources
  if (!ValidateConstWhileParams(*params, status)) return;
  TF_FinishWhileHelper(params, status, outputs);
  FreeWhileResources(params);