memset(buffer, 'x', n);
    TF_SetStatus(status, TF_OK, "");
    return n;
  };
  const uint32 block_count = 256;
  tf_gcs_filesystem::RamFileBlockCache cache(
      block_size, block_count * block_size, 0, fetcher);
  std::vector<char> out;
  out.resize(block_count, 0);
  // The cache has space for `block_count` blocks. The loop with i = 0 should
  // fill the cache, and the loop with i = 1 should be all cache hits. The

void TF_DeleteRecursively(const char* dirname, int64_t* undeleted_file_count,
                          int64_t* undeleted_dir_count, TF_Status* status) {
  ::int64_t f, d;

  TF_SetStatus(status, TF_OK, "");
  ::tensorflow::Set_TF_Status_from_Status(
      status, ::tensorflow::Env::Default()->DeleteRecursively(dirname, &f, &d));
  *undeleted_file_count = f;
  *undeleted_dir_count = d;
}

# proto-message: ActionSpecs

# Python rules should not have more than one source file.
action_spec {
  action: CHECK_FILE_COUNT
  file_count_params {
    rule_selector {
      rule_kind_regex: "^.*py(type)?(_strict)?_(binary|library|test).*$"
      generator_function_regex: "^(?!boq_header)$"
    }
    max_source_count: 1
  }

// set to 1, *undeleted_file_count is set to 0 and an appropriate status (e.g.
// TF_NOT_FOUND) is returned.
//
// If dirname and all its descendants were successfully deleted, TF_OK is
// returned and both error counters are set to zero.
//
// Otherwise, while traversing the tree, undeleted_file_count and
// undeleted_dir_count are updated if an entry of the corresponding type could

                    verbose: bool = False):
  """Creates a JUnit-based XML file, with each invocation as a testcase."""
  os.makedirs(os.path.dirname(output_path), exist_ok=True)
  failure_count = 0
  error_count = 0

  date_time = datetime.datetime
  attrib = {'name': 'Bazel Invocations', 'time': '0.0',
            'timestamp': date_time.isoformat(date_time.utcnow())}
  testsuites = ElemTree.Element('testsuites')

                                                           int bucket_count) {
  return new TFE_MonitoringBuckets([scale, growth_factor, bucket_count]() {
    return tensorflow::monitoring::Buckets::Exponential(scale, growth_factor,
                                                        bucket_count);
  });
}

void TFE_MonitoringDeleteBuckets(TFE_MonitoringBuckets* buckets) {
  delete buckets;

          }
        }
      }
      auto usage_count_it = state.tensor_usage_counts.find(id);
      if (usage_count_it == state.tensor_usage_counts.end()) {
        VLOG(1) << "Tensor " << id << " not used";
        continue;
      }
      usage_count_it->second--;
      if (usage_count_it->second > 0) {
        VLOG(1) << "Tensor " << id << " usage count " << usage_count_it->second;
        continue;
      }

                   TF_Message(status));
    }

    if (device_index == 0) {
      first_op_output_count = per_device_output_tensors.rbegin()->size();
    } else {
      if (first_bad_status == nullptr &&
          per_device_output_tensors.rbegin()->size() != first_op_output_count) {
        first_bad_status.reset(TF_NewStatus());
        TF_SetStatus(first_bad_status.get(), TF_INTERNAL,

      new TensorSliceReader::VarToDataTypeMap);
  v2_reader_->Seek(kHeaderEntryKey);
  for (v2_reader_->Next(); v2_reader_->Valid(); v2_reader_->Next()) {
    if (filtered_keys.count(string(v2_reader_->key())) > 0) continue;
    CHECK(entry.ParseFromArray(v2_reader_->value().data(),
                               v2_reader_->value().size()))
        << entry.InitializationErrorString();

    if (downloaded_block) {
      absl::MutexLock l(&mu_);
      // Do not update state if the block is already to be evicted.
      if (block->timestamp != 0) {
        // Use capacity() instead of size() to account for all  memory
        // used by the cache.
        cache_size_ += block->data.capacity();
        // Put to beginning of LRA list.
        lra_list_.erase(block->lra_iterator);
        lra_list_.push_front(key);