CVE-2021-37646
Description
TensorFlow is an end-to-end open source platform for machine learning. In affected versions the implementation of `tf.raw_ops.StringNGrams` is vulnerable to an integer overflow issue caused by converting a signed integer value to an unsigned one and then allocating memory based on this value. The [implementation](https://github.com/tensorflow/tensorflow/blob/8d72537c6abf5a44103b57b9c2e22c14f5f49698/tensorflow/core/kernels/string_ngrams_op.cc#L184) calls `reserve` on a `tstring` with a value that sometimes can be negative if user supplies negative `ngram_widths`. The `reserve` method calls `TF_TString_Reserve` which has an `unsigned long` argument for the size of the buffer. Hence, the implicit conversion transforms the negative value to a large integer. We have patched the issue in GitHub commit c283e542a3f422420cfdb332414543b62fc4e4a5. The fix will be included in TensorFlow 2.6.0. We will also cherrypick this commit on TensorFlow 2.5.1, TensorFlow 2.4.3, and TensorFlow 2.3.4, as these are also affected and still in supported range.
How to fix
- google tensorflow2.3.4, 2.4.3 or later
- tensorflow2.3.4, 2.4.3, 2.5.1 or later
- tensorflow-cpu2.3.4, 2.4.3, 2.5.1 or later
- tensorflow-gpu2.3.4, 2.4.3, 2.5.1 or later
Affected products
| Product | Affected versions | Fixed in |
|---|---|---|
| google tensorflow | 2.3.0 ≤ v < 2.3.4, 2.4.0 ≤ v < 2.4.3, 2.5.0, 2.6.0 | 2.3.4, 2.4.3 |
| tensorflowPyPI | v < 2.3.4, 2.4.0 ≤ v < 2.4.3, 2.5.0 ≤ v < 2.5.1, 2.3.0 ≤ v < 2.3.4 | 2.3.4, 2.4.3, 2.5.1 |
| tensorflow-cpuPyPI | v < 2.3.4, 2.4.0 ≤ v < 2.4.3, 2.5.0 ≤ v < 2.5.1, 2.3.0 ≤ v < 2.3.4 | 2.3.4, 2.4.3, 2.5.1 |
| tensorflow-gpuPyPI | v < 2.3.4, 2.4.0 ≤ v < 2.4.3, 2.5.0 ≤ v < 2.5.1, 2.3.0 ≤ v < 2.3.4 | 2.3.4, 2.4.3, 2.5.1 |
Check your version
Choose the product and enter the version you run to see whether it falls in the affected range.
Attack Surface Management
See the assets tied to your domains, the technologies they run and known vulnerabilities in one list.