// Get a list of the names of functions that have been registered.
  virtual std::vector<string> ListFunctionNames() = 0;

  struct CacheStats {
    int64_t kernel_cache_size;
    int64_t device_cache_size;
    std::map<std::string, int64_t> func_kernel_cache_entries;
    int64_t local_rendezvous_cache_active_size;
  };
  virtual CacheStats GetCacheStats() = 0;

TF_CAPI_EXPORT extern bool TFE_ExecutorIsAsync(TFE_Executor*);

// Causes the calling thread to block till all ops dispatched in this executor
// have been executed. Note that "execution" here refers to kernel execution /
// scheduling of copies, etc. Similar to sync execution, it doesn't guarantee
// that lower level device queues (like GPU streams) have been flushed.
//

        based server to register tf.function methods and a GRPC client to invoke
        remote registered methods. RPC APIs are intended for multi-client setups
        i.e. server and clients are started in separate binaries independently.

    *   Example usage to create server: ```python server =
        tf.distribute.experimental.rpc.Server.create("grpc", "127.0.0.1:1234")

#     visibility = ["//visibility:public"],
# )
# copybara:uncomment_end

# 'enable_registration_v2' opts-in to a different implementation of op and
# kernel registration - REGISTER_OP, REGISTER_KERNEL_BUILDER, etc.
#
# This setting is currently experimental. The 'v2' implementation does _not_
# correspond to a particular, finalized design; rather, it relates to
# developing one.
#

        "//tensorflow/core/distributed_runtime/rpc:grpc_server_lib",
        "@com_google_absl//absl/strings",
    ],
)

tf_cuda_cc_test(
    name = "c_api_remote_test",
    size = "medium",
    srcs = [
        "c_api_remote_test.cc",
    ],
    # TODO(b/136478427): Figure out how to correctly shut the server down
    args = ["--heap_check="],
    tags = [
        "no_windows",

///     break ABI compatibility and cause all plugins to be recompiled, these
///     type of changes should be extremely rare.
///
/// Next section will detail this as well as some corner cases that are out of
/// scope for now.

// LINT.IfChange
typedef struct TF_RandomAccessFileOps {
  /// Releases resources associated with `*file`.
  ///

build:mkl --define=tensorflow_mkldnn_contraction_kernel=0
build:mkl --define=build_with_openmp=true
build:mkl -c opt

# config to build OneDNN backend with a user specified threadpool.
build:mkl_threadpool --define=build_with_mkl=true --define=enable_mkl=true
build:mkl_threadpool --define=tensorflow_mkldnn_contraction_kernel=0
build:mkl_threadpool --define=build_with_mkl_opensource=true

  VLOG(1) << "Enqueuing is done.";
}

TF_Buffer* TFE_GetServerDef(const char* text_proto, TF_Status* status) {
  tensorflow::ServerDef server_def;
  if (!tensorflow::protobuf::TextFormat::ParseFromString(text_proto,
                                                         &server_def)) {
    status->status = tensorflow::errors::Internal(
        "Invalid text proto for ServerDef: ", text_proto);
    return nullptr;
  }

# limitations under the License.
# ==============================================================================
setup_file() {
    bazel version  # Start the bazel server
}

# Do a bazel query specifically for the licenses checker. It searches for
# targets matching the provided query, which start with // or @ but not
# //tensorflow (so it looks for //third_party, //external, etc.), and then

  tensorflow::ServerDef server_def = GetServerDef(2);

  // This server def has the task index set to 0.
  string serialized = server_def.SerializeAsString();

  server_def.set_task_index(1);

  std::unique_ptr<tensorflow::GrpcServer> worker_server;
  ASSERT_TRUE(tensorflow::GrpcServer::Create(
                  server_def, tensorflow::Env::Default(), &worker_server)
                  .ok());