A vulnerability was found in TensorFlow up to 2.2.0/2.3.0 (Artificial Intelligence Software). It has been declared as critical. This vulnerability affects the function dlpack.to_dlpack. The manipulation as part of a Argument leads to a uninitialized resource vulnerability. The CWE definition for the vulnerability is CWE-908. The product uses or accesses a resource that has not been initialized. As an impact it is known to affect confidentiality, integrity, and availability. CVE summarizes:

In Tensorflow before versions 2.2.1 and 2.3.1, 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. However, there is nothing stopping users from passing in a Python object instead of a tensor. The uninitialized memory address is due to a `reinterpret_cast` Since the `PyObject` is a Python object, not a TensorFlow Tensor, the cast to `EagerTensor` fails. The issue is patched in commit 22e07fb204386768e5bcbea563641ea11f96ceb8 and is released in TensorFlow versions 2.2.1, or 2.3.1.

The weakness was presented 09/25/2020 (GitHub Repository). The advisory is available at github.com. This vulnerability was named CVE-2020-15193 since 06/25/2020. The attack can be initiated remotely. A single authentication is needed for exploitation. Technical details are known, but there is no available exploit.

Upgrading to version 2.2.1 or 2.3.1 eliminates this vulnerability.

See 162002, 162001, 162000 and 161999 for similar entries.

Productinfo

Type

• Artificial Intelligence Software

Name

• TensorFlow

Version

• 2.0
• 2.1
• 2.2
• 2.3

License

• free

CPE 2.3info

• 🔍
• 🔍
• 🔍

CPE 2.2info

• 🔍
• 🔍
• 🔍

CVSSv4info

CVSSv3info

CVSSv2info

Exploitinginfo

Threat Intelligenceinfo

Countermeasuresinfo

Timelineinfo

Sourcesinfo

Entryinfo

Discussion