if [[ -n "$(which grealpath)" ]] &&  [[ -n "$(which gstat)" ]]; then
  alias realpath=grealpath
  alias stat=gstat
  # By default, aliases are only expanded in interactive shells, which means
  # that they are not substituted for their corresponding commands in shell
  # scripts. By setting "expand_aliases", we enable alias expansion in
  # non-interactive shells as well.
  shopt -s expand_aliases
else

@@ -125,14 +131,16 @@
    If weak aliases are not available, this defines a strong alias.  */
 # define weak_alias(name, aliasname) _weak_alias (name, aliasname)
 # define _weak_alias(name, aliasname) \
-  extern __typeof (name) aliasname __attribute__ ((weak, alias (#name)));
+  extern __typeof (name) aliasname __attribute__ ((weak, alias (#name))) \
+    __attribute_copy__ (name);
 

    int64_t dims[] = {2, 2};
    TFE_TensorHandle* copy_aliased = TFE_NewTensorHandleFromDeviceMemory(
        ctx, name, TF_FLOAT, dims, 2, data, size, &Deleter, copy, status);
    CHECK_EQ(TF_OK, TF_GetCode(status)) << TF_Message(status);
    TFE_TensorHandle* on_host =
        TFE_TensorHandleCopyToDevice(copy_aliased, ctx, "CPU:0", status);
    CHECK_EQ(TF_OK, TF_GetCode(status)) << TF_Message(status);

if _os.getenv("TF_PLUGGABLE_DEVICE_LIBRARY_PATH", ""):
  _ll.load_pluggable_device_library(
      _os.getenv("TF_PLUGGABLE_DEVICE_LIBRARY_PATH")
  )

# Add Keras module aliases
_losses = _KerasLazyLoader(globals(), submodule="losses", name="losses")
_metrics = _KerasLazyLoader(globals(), submodule="metrics", name="metrics")
_optimizers = _KerasLazyLoader(

  # version.
  meta = open(os.path.join(android_ndk_home_path, 'meta/platforms.json'))
  platforms = json.load(meta)
  meta.close()
  aliases = platforms['aliases']
  api_levels = sorted(list(set([aliases[i] for i in aliases])))

  android_ndk_api_level = prompt_loop_or_load_from_env(
      environ_cp,
      var_name='ANDROID_NDK_API_LEVEL',

_current_module.app.flags = flags  # pylint: disable=undefined-variable
setattr(_current_module, "flags", flags)

_major_api_version = 1

# Add module aliases from Keras to TF.
# Some tf endpoints actually lives under Keras.
_current_module.layers = _KerasLazyLoader(
    globals(),
    submodule="__internal__.legacy.layers",
    name="layers",
    mode="v1")

[ -z "$PS1" ] && return

# Set up attractive prompt
export PS1="\[\e[31m\]tf-docker\[\e[m\] \[\e[33m\]\w\[\e[m\] > "
export TERM=xterm-256color
alias grep="grep --color=auto"
alias ls="ls --color=auto"
# Fix nvidia-docker

TF_CAPI_EXPORT extern TFE_TensorHandle* TFE_TensorHandleCopySharingTensor(
    TFE_TensorHandle* h, TF_Status* status);

// This function will block till the operation that produces `h` has
// completed. The memory returned might alias the internal memory used by
// TensorFlow. Hence, callers should not mutate this memory (for example by
// modifying the memory region pointed to by TF_TensorData() on the returned
// TF_Tensor).

    * Introduced API `tf.saved_model.experimental.read_fingerprint(export_dir)` for reading the fingerprint of a SavedModel.
* `tf.random`
  * Added non-experimental aliases for `tf.random.split` and `tf.random.fold_in`, the experimental endpoints are still available so no code changes are necessary.
* `tf.experimental.ExtensionType`

alias(
    name = "is_cuda_enabled",
    actual = if_oss(
        "@local_config_cuda//:is_cuda_enabled",
        "@local_config_cuda//cuda:using_clang",
    ),
)

# Config setting that is satisfied when CUDA device code should be compiled
# with clang. It does not imply that CUDA support has been enabled.
alias(
    name = "is_cuda_compiler_clang",