Invalid validation in `QuantizeAndDequantizeV2`
Low3.6CVE-2021-29610 · Published May 21, 2021 · updated Jul 8, 2026
Affected versions
| Package | Affected | Fixed in |
|---|---|---|
| tensorflow PyPI | < 2.1.4 | 2.1.4 |
| >= 2.2.0, < 2.2.3 | 2.2.3 | |
| >= 2.3.0, < 2.3.3 | 2.3.3 | |
| >= 2.4.0, < 2.4.2 | 2.4.2 |
Details and references
### Impact The validation in `tf.raw_ops.QuantizeAndDequantizeV2` allows invalid values for `axis` argument: ```python import tensorflow as tf input_tensor = tf.constant([0.0], shape=[1], dtype=float) input_min = tf.constant(-10.0) input_max = tf.constant(-10.0) tf.raw_ops.QuantizeAndDequantizeV2( input=input_tensor, input_min=input_min, input_max=input_max, signed_input=False, num_bits=1, range_given=False, round_mode='HALF_TO_EVEN', narrow_range=False, axis=-2) ``` The [validation](https://github.com/tensorflow/tensorflow/blob/eccb7ec454e6617738554a255d77f08e60ee0808/tensorflow/core/kernels/quantize_and_dequantize_op.cc#L74-L77) uses `||` to mix two different conditions: ```cc OP_REQUIRES(ctx, (axis_ == -1 || axis_ < input.shape().dims()), errors::InvalidArgument(...)); ``` If `axis_ < -1` the condition in `OP_REQUIRES` will still be true, but this value of `axis_` results in heap underflow. This allows attackers to read/write to other data on the heap. ### Patches We have patched the issue in GitHub commit [c5b0d5f8ac19888e46ca14b0e27562e7fbbee9a9](https://github.com/tensorflow/tensorflow/commit/c5b0d5f8ac19888e46ca14b0e27562e7fbbee9a9). The fix will be included in TensorFlow 2.5.0. We will also cherrypick this commit on TensorFlow 2.4.2, TensorFlow 2.3.3, TensorFlow 2.2.3 and TensorFlow 2.1.4, as these are also affected and still in supported range. ### 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 Yakun Zhang and Ying Wang of Baidu X-Team.
- CVSS 3.1
- CVSS:3.1/AV:L/AC:H/PR:L/UI:N/S:U/C:N/I:L/A:L
- Severity from
- GitHub (reviewed advisory)
- Weakness
- CWE-665, CWE-787
- Also known as
- BIT-tensorflow-2021-29610, CVE-2021-29610, PYSEC-2021-247, PYSEC-2021-538, PYSEC-2021-736
- github.com/tensorflow/tensorflow/security/advisories/GHSA-mq5c-prh3-3f3h
- nvd.nist.gov/vuln/detail/CVE-2021-29610
- github.com/tensorflow/tensorflow/commit/c5b0d5f8ac19888e46ca14b0e27562e7fbbee9a9
- github.com/pypa/advisory-database/tree/main/vulns/tensorflow-cpu/PYSEC-2021-538.yaml
- github.com/pypa/advisory-database/tree/main/vulns/tensorflow-gpu/PYSEC-2021-736.yaml
- github.com/pypa/advisory-database/tree/main/vulns/tensorflow/PYSEC-2021-247.yaml
- github.com/tensorflow/tensorflow
More TensorFlow advisories
All TensorFlow| Date | Advisory | Severity | Fixed in |
|---|---|---|---|
| May 212021 | Heap buffer overflow in `RaggedBinCount` CVE-2021-29512Low2.5fixed in 2.3.3, 2.4.2 | Low2.5 | 2.3.3, 2.4.2 |
| May 212021 | Type confusion during tensor casts lead to dereferencing null pointers CVE-2021-29513Low2.5fixed in 2.1.4, 2.2.3, 2.3.3, 2.4.2 | Low2.5 | 2.1.4, 2.2.3, 2.3.3, 2.4.2 |
| May 212021 | Heap out of bounds write in `RaggedBinCount` CVE-2021-29514Low2.5fixed in 2.3.3, 2.4.2 | Low2.5 | 2.3.3, 2.4.2 |
| May 212021 | Reference binding to null pointer in `MatrixDiag*` ops CVE-2021-29515Low2.5fixed in 2.1.4, 2.2.3, 2.3.3, 2.4.2 | Low2.5 | 2.1.4, 2.2.3, 2.3.3, 2.4.2 |
| May 212021 | Null pointer dereference via invalid Ragged Tensors CVE-2021-29516Low2.5fixed in 2.1.4, 2.2.3, 2.3.3, 2.4.2 | Low2.5 | 2.1.4, 2.2.3, 2.3.3, 2.4.2 |
| May 212021 | Division by zero in `Conv3D` CVE-2021-29517Low2.5fixed in 2.1.4, 2.2.3, 2.3.3, 2.4.2 | Low2.5 | 2.1.4, 2.2.3, 2.3.3, 2.4.2 |