[],
        otherwise = [
            "@local_xla//xla/stream_executor/cuda:all_runtime",
            "@local_xla//xla/stream_executor/cuda:cuda_platform",
            "@local_xla//xla/stream_executor/cuda:cudnn_plugin",
            "@local_xla//xla/stream_executor/cuda:cufft_plugin",
            "@local_xla//xla/stream_executor:cuda_platform",
        ],
    ),
    deps = if_cuda([

        "//tensorflow/core/transforms:__subpackages__",
    ],
    deps = [
        ":c_api_macros_hdrs",
        "@local_xla//xla/tsl/c:tsl_status",
        "@local_xla//xla/tsl/c:tsl_status_internal",
        "@local_xla//xla/tsl/platform:status",
    ] + select({
        "//tensorflow:android": [
            "//tensorflow/core:portable_tensorflow_lib_lite",  # TODO(annarev): exclude runtime srcs

WIP ML Build Docker container for ML repositories (Tensorflow, JAX and XLA).

This container branches off from
/tensorflow/tools/tf_sig_build_dockerfiles/. However, since
hermetic CUDA and hermetic Python is now available for Tensorflow, a lot of the
requirements installed on the original container can be removed to reduce the
footprint of the container and make it more reusable across different ML

  cp tensorflow/core/platform/ctstring.h \
    tensorflow/core/platform/ctstring_internal.h \
    ${LIB_PKG}/include/tensorflow/core/platform
  cp third_party/xla/xla/tsl/c/tsl_status.h ${LIB_PKG}/include/xla/tsl/c
  cp third_party/xla/third_party/tsl/tsl/platform/ctstring.h \
     third_party/xla/third_party/tsl/tsl/platform/ctstring_internal.h \
     ${LIB_PKG}/include/tsl/platform
  cd ${LIB_PKG}

#     --config=dbg --per_file_copt=+tensorflow/core/kernels/identity_op.*@-g
# Since this .bazelrc file is synced between the tensorflow/tensorflow repo and
# the openxla/xla repo, also include debug info for files under xla/.
build:dbg --per_file_copt=+.*,-tensorflow.*,-xla.*@-g0
build:dbg --per_file_copt=+tensorflow/core/kernels.*@-g0
# for now, disable arm_neon. see: https://github.com/tensorflow/tensorflow/issues/33360

**Linux XLA**                 | [![Status](https://storage.googleapis.com/tensorflow-kokoro-build-badges/ubuntu-xla.svg)](https://storage.googleapis.com/tensorflow-kokoro-build-badges/ubuntu-xla.html)         | TBA

  auto* optimizer_options =
      config.mutable_graph_options()->mutable_optimizer_options();
  if (enable) {
    optimizer_options->set_global_jit_level(tensorflow::OptimizerOptions::ON_1);

    // These XLA flags are needed to trigger XLA properly from C (more generally
    // non-Python) clients. If this API is called again with `enable` set to
    // false, it is safe to keep these flag values as is.
    tensorflow::MarkForCompilationPassFlags* flags =

reachable and exploitable through production-grade, benign models.

### Compilation

Compiling models via the recommended entry points described in
[XLA](https://www.tensorflow.org/xla) and
[JAX](https://jax.readthedocs.io/en/latest/jax-101/02-jitting.html)
documentation should be safe, while some of the testing and debugging tools that

        Eager mode.

*   `tf.lite`:

    *   Enable TFLite experimental new converter by default.

*   XLA

    *   XLA now builds and works on windows. All prebuilt packages come with XLA
        available.
    *   XLA can be
        [enabled for a `tf.function`](https://www.tensorflow.org/xla#explicit_compilation_with_tffunction)
        with “compile or throw exception” semantics on CPU and GPU.

                ' of compute capabilities excluding version %s.' % ver)
            all_valid = False
          if ver < 35:
            print('WARNING: XLA does not support CUDA compute capabilities '
                  'lower than sm_35. Disable XLA when running on older GPUs.')
      else:
        ver = float(m.group(0))
        if ver < 3.0: