## TFSA-2020-023: Memory corruption in `dlpack.to_dlpack` ### CVE Number CVE-2020-15193 ### Impact The implementation of `dlpack.to_dlpack` can be made to use uninitialized memory resulting in further memory corruption. This is because the pybind11 glue code [assumes that the argument is a tensor](https://github.com/tensorflow/tensorflow/blob/0e68f4d3295eb0281a517c3662f6698992b7b2cf/tensorflow/python/tfe_wrapper.cc#L1361): ```cc TFE_TensorHandle* thandle = EagerTensor_Handle(eager_tensor_pyobject_ptr); ``` However, there is nothing stopping users from passing in a Python object instead of a tensor. ```python In [2]: tf.experimental.dlpack.to_dlpack([2]) ==1720623==WARNING: MemorySanitizer: use-of-uninitialized-value #0 0x55b0ba5c410a in tensorflow::(anonymous namespace)::GetTensorFromHandle(TFE_TensorHandle*, TF_Status*) third_party/tensorflow/c/eager/dlpack.cc:46:7 #1 0x55b0ba5c38f4 in tensorflow::TFE_HandleToDLPack(TFE_TensorHandle*, TF_Status*) third_party/tensorflow/c/eager/dlpack.cc:252:26 ... ``` The uninitialized memory address is due to a [`reinterpret_cast`](https://github.com/tensorflow/tensorflow/blob/0e68f4d3295eb0281a517c3662f6698992b7b2cf/tensorflow/python/eager/pywrap_tensor.cc#L848-L850): ```cc TFE_TensorHandle* EagerTensor_Handle(const PyObject* o) { return reinterpret_cast<const EagerTensor*>(o)->handle; } ``` Since the `PyObject` is a Python object, not a TensorFlow Tensor, the cast to `EagerTensor` fails. ### Vulnerable Versions TensorFlow 2.2.0, 2.3.0. ### Patches We have patched the issue in [22e07fb204386768e5bcbea563641ea11f96ceb8](https://github.com/tensorflow/tensorflow/commit/22e07fb204386768e5bcbea563641ea11f96ceb8) and will release a patch release for all affected versions. We recommend users to upgrade to TensorFlow 2.2.1 or 2.3.1. ### For more information Please consult [our security guide](https://github.com/tensorflow/tensorflow/blob/master/SECURITY.md) 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.