break;
      case FetchState::FINISHED:
        return TF_SetStatus(status, TF_OK, "");
    }
  }
  return TF_SetStatus(
      status, TF_INTERNAL,
      "Control flow should never reach the end of RamFileBlockCache::Fetch.");
}

int64_t RamFileBlockCache::Read(const std::string& filename, size_t offset,
                                size_t n, char* buffer, TF_Status* status) {
  if (n == 0) {

  // fill the cache, and the loop with i = 1 should be all cache hits. The
  // fetcher checks that it is called once and only once for each offset (to
  // fetch the corresponding block).
  for (int i = 0; i < 2; i++) {
    for (int j = 0; j < block_count; j++) {
      TF_EXPECT_OK(ReadCache(&cache, "", block_size * j, block_size, &out));
    }
  }
}

typedef enum TF_TriState {
  TF_TriState_Default = 0,
  TF_TriState_Off,
  TF_TriState_On,
} TF_TriState;

// TF_GrapplerItem represents a combination of a graph, one of more fetch nodes,
// and potentially a set of nodes to feed.
typedef struct TF_GrapplerItem TF_GrapplerItem;

// Flags indicating whether existing optimizers should be turned off.

                              size_t* storage_size, TF_Status* status) {
  TF_SetStatus(status, TF_OK, "");
  const std::vector<std::string>& nodes =
      reinterpret_cast<const tensorflow::grappler::GrapplerItem*>(item)->fetch;
  *num_values = nodes.size();
  *storage_size = 0;
  for (const std::string& str : nodes) {
    *storage_size += str.size();
  }
}

void TF_GetFetchNodesList(const TF_GrapplerItem* item, char** values,

        `pip install -U tf-nightly`
    *   Rebase your pull request onto a recent pull from
        [TensorFlow's](https://github.com/tensorflow/tensorflow) master branch.

2.  If you are changing the code and the docstring of a class/function/method,
    then you will need to
    [build TensorFlow from source](https://www.tensorflow.org/install/source).

// inspect values. This would let people e.g. copy over most attributes and then
// modify some based on their values.

// A reference to an op's name -> attribute mapping
typedef struct TFE_OpAttrs TFE_OpAttrs;

// Fetch a reference to `op`'s attributes. The returned reference is only valid
// while `op` is alive.
TF_CAPI_EXPORT extern const TFE_OpAttrs* TFE_OpGetAttrs(const TFE_Op* op);
// Add attributes in `attrs` to `op`.
//

# Workaround for networking (b/112379377) was closed as won't fix for MTU setting.
# Remaining lines handle making the metadata server on the VM accessible inside docker.
RUN Get-NetAdapter | Where-Object Name -like "*Ethernet*" | ForEach-Object { \
        & netsh interface ipv4 set subinterface $_.InterfaceIndex mtu=1460 store=persistent }; \
    $gateway = (Get-NetRoute | Where { $_.DestinationPrefix -eq \"0.0.0.0/0\" } | Sort-Object RouteMetric \

// run on device_type.
TF_CAPI_EXPORT extern void TF_AttrBuilderCheckCanRunOnDevice(
    TF_AttrBuilder* builder, const char* device_type, TF_Status* status);

// For argument number input_index, fetch the corresponding number_attr that
// needs to be updated with the argument length of the input list.
// Returns nullptr if there is any problem like op_name is not found, or the
// argument does not support this attribute type.

  EXPECT_EQ(item.fetch.size(), num_values);
  EXPECT_EQ(list_total_size, storage_size);

  std::unique_ptr<char*[]> values(new char*[item.fetch.size()]);
  std::unique_ptr<size_t[]> lens(new size_t[item.fetch.size()]);
  std::unique_ptr<char[]> storage(new char[storage_size]);
  TF_GetFetchNodesList(c_item, values.get(), lens.get(), item.fetch.size(),

Follow these steps to patch a specific version of TensorFlow, for example, to
apply fixes to bugs or security vulnerabilities:

*   Clone the TensorFlow repo and switch to the corresponding branch for your
    desired TensorFlow version, for example, branch `r2.8` for version 2.8.
*   Apply (that is, cherry-pick) the desired changes and resolve any code
    conflicts.
*   Run TensorFlow tests and ensure they pass.