tfsa-2020-005.md

TFSA-2020-005: Out of bounds access in TFLite operators

CVE Number

CVE-2020-15211

Impact

In TensorFlow Lite, saved models in the flatbuffer format use a double indexing scheme: a model has a set of subgraphs, each subgraph has a set of operators and each operator has a set of input/output tensors. The flatbuffer format uses indices for the tensors, indexing into an array of tensors that is owned by the subgraph. This results in a pattern of double array indexing when trying to get the data of each tensor:

  return &context->tensors[node->inputs->data[index]];

However, some operators can have some tensors be optional. To handle this scenario, the flatbuffer model uses a negative -1 value as index for these tensors:

#define kTfLiteOptionalTensor (-1)

This results in special casing during validation at model loading time:

  for (int i = 0; i < length; i++) {
    int index = indices[i];
    // Continue if index == kTfLiteOptionalTensor before additional comparisons
    // below, size_t(-1) is always >= context_tensors_size.
    if (index == kTfLiteOptionalTensor) {
      continue;
    }
    if (index < 0 || static_cast<size_t>(index) >= context_.tensors_size) {
      ReportError(
          "Invalid tensor index %d in %s. The subgraph has %d tensors\n", index,
          label, context_.tensors_size);
      consistent_ = false;
      return kTfLiteError;
    }
  }

Unfortunately, this means that the -1 index is a valid tensor index for any operator, including those that don't expect optional inputs and including for output tensors. Thus, this allows writing and reading from outside the bounds of heap allocated arrays, although only at a specific offset from the start of these arrays.

This results in both read and write gadgets, albeit very limited in scope.

Vulnerable Versions

TensorFlow 1.15.0, 1.15.1, 1.15.2, 1.15.3, 2.0.0, 2.0.1, 2.0.2, 2.1.0, 2.1.1, 2.2.0, 2.3.0.

Patches

We have patched the issue in several commits (46d5b0852, 00302787b7, e11f5558, cd31fd0ce, 1970c21, and fff2c83). We will release patch releases for all versions between 1.15 and 2.3.

We recommend users to upgrade to TensorFlow 1.15.4, 2.0.3, 2.1.2, 2.2.1, or 2.3.1.

Workarounds

A potential workaround would be to add a custom Verifier to the model loading code to ensure that only operators which accept optional inputs use the -1 special value and only for the tensors that they expect to be optional. Since this allow-list type approach is erro-prone, we advise upgrading to the patched code.

For more information

Please consult our security guide for more information regarding the security model and how to contact us with issues and questions.

Attribution

This vulnerability has been reported by members of the Aivul Team from Qihoo 360.