# copybara:uncomment_begin(google-only)
            # "-Wl,--undefined-version",
            # copybara:uncomment_end(google-only)
        ],
    }),
    linkstatic = 1,
    per_os_targets = True,
    roots = [
                "//tensorflow/c/experimental/filesystem:filesystem_interface",
                "//tensorflow/c/experimental/stream_executor:stream_executor",

* <SIMILAR TO ABOVE SECTION, BUT FOR OTHER IMPORTANT CHANGES / BUG FIXES>
* <IF A CHANGE CLOSES A GITHUB ISSUE, IT SHOULD BE DOCUMENTED HERE>
* <NOTES SHOULD BE GROUPED PER AREA>

* `tf.lite`
    * Quantization for `FullyConnected` layer is switched from per-tensor to
      per-channel scales for dynamic range quantization use case (`float32`
      inputs / outputs and `int8` weights). The change enables new quantization

          exempt-pr-labels: "override-stale"
          #Limit the No. of API calls in one run default value is 30.
          operations-per-run: 1000
          days-before-issue-stale: 7
          days-before-issue-close: 7
          stale-issue-label: "stale"
          # reason for closed the issue default value is not_planned

    licenses = ["notice"],
)

# Filesystem implementation for GCS environments
tf_cc_shared_object(
    name = "gcs_filesystem",
    framework_so = [],
    linkstatic = False,
    per_os_targets = 1,
    visibility = ["//visibility:public"],
    deps = [":gcs_filesystem_impl"],
)

# The real implementation of the filesystem.
cc_library(
    name = "gcs_filesystem_impl",

# Users can still include debug info for a specific kernel, e.g. with:
#     --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

  // once) and produces the gradient of the target tensors with respect to the
  // source tensors. The output gradients are used if not empty and not
  // null. The result is populated with one tensor per target element.
  // When running backward functions, builds zeros-like tensors for
  // incoming grads which are nullptrs, unless `build_default_zeros_grads`
  // is set to false.
  Status ComputeGradient(

    TF_Status* status) const {
  absl::optional<std::vector<std::unique_ptr<ParallelTensor>>> result;
  // Compute per-device per-output tensors
  std::vector<std::vector<TensorHandlePtr>> per_device_output_tensors;
  per_device_output_tensors.reserve(underlying_devices_.size());
  int first_op_output_count = 0;
  StatusPtr first_bad_status(nullptr);

  const char task1_name[] = "/job:localhost/replica:0/task:1/device:CPU:0";
  const char task2_name[] = "/job:localhost/replica:0/task:2/device:CPU:0";

  // Create one variable per task.
  TFE_TensorHandle* h0 = TestVariable(ctx, 1.0, task1_name);
  TFE_TensorHandle* h1 = TestVariable(ctx, 2.0, task2_name);
  TFE_TensorHandle* h2 = TestVariable(ctx, 3.0, task0_name);