{"id":"CVE-2020-15265","aliases":["GHSA-rrfp-j2mp-hq9c","BIT-tensorflow-2020-15265","PYSEC-2020-138","PYSEC-2020-295","PYSEC-2020-330"],"title":"Segfault in `tf.quantization.quantize_and_dequantize`","summary":"Segfault in `tf.quantization.quantize_and_dequantize`","severity":"medium","cvss":5.9,"cvssVector":"CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:N/I:N/A:H","vendor":"tensorflow","product":"tensorflow","ecosystem":"pip","affected":["tensorflow < 2.4.0","tensorflow-cpu < 2.4.0","tensorflow-gpu < 2.4.0"],"patched":["tensorflow 2.4.0","tensorflow-cpu 2.4.0","tensorflow-gpu 2.4.0"],"published":"2020-11-13","updated":"2026-07-08","source":"OSV","sourceUrl":"https://osv.dev/vulnerability/GHSA-rrfp-j2mp-hq9c","references":[{"url":"https://github.com/tensorflow/tensorflow/security/advisories/GHSA-rrfp-j2mp-hq9c"},{"url":"https://nvd.nist.gov/vuln/detail/CVE-2020-15265"},{"url":"https://github.com/tensorflow/tensorflow/issues/42105"},{"url":"https://github.com/tensorflow/tensorflow/commit/eccb7ec454e6617738554a255d77f08e60ee0808"},{"url":"https://github.com/pypa/advisory-database/tree/main/vulns/tensorflow-cpu/PYSEC-2020-295.yaml"},{"url":"https://github.com/pypa/advisory-database/tree/main/vulns/tensorflow-gpu/PYSEC-2020-330.yaml"},{"url":"https://github.com/pypa/advisory-database/tree/main/vulns/tensorflow/PYSEC-2020-138.yaml"},{"url":"https://github.com/tensorflow/tensorflow"}],"tags":["osv","pip"],"epss":0.00908,"epssPercentile":0.58451,"ingestedAt":"2026-07-08T18:25:53.013Z","slug":"CVE-2020-15265","body":"## Overview\n\n### Impact\nAn attacker can pass an invalid `axis` value to `tf.quantization.quantize_and_dequantize`:\n\n```python\ntf.quantization.quantize_and_dequantize(\n    input=[2.5, 2.5], input_min=[0,0], input_max=[1,1], axis=10)\n```\n\nThis results in accessing [a dimension outside the rank of the input tensor](https://github.com/tensorflow/tensorflow/blob/0225022b725993bfc19b87a02a2faaad9a53bc17/tensorflow/core/kernels/quantize_and_dequantize_op.cc#L74) in the C++ kernel implementation:\n```\nconst int depth = (axis_ == -1) ? 1 : input.dim_size(axis_);\n```\n\nHowever, [`dim_size` only does a `DCHECK`](https://github.com/tensorflow/tensorflow/blob/0225022b725993bfc19b87a02a2faaad9a53bc17/tensorflow/core/framework/tensor_shape.cc#L292-L307) to validate the argument and then uses it to access the corresponding element of an array:\n```\nint64 TensorShapeBase<Shape>::dim_size(int d) const {\n  DCHECK_GE(d, 0);\n  DCHECK_LT(d, dims());\n  DoStuffWith(dims_[d]);\n}\n```\n\nSince in normal builds, `DCHECK`-like macros are no-ops, this results in segfault and access out of bounds of the array.\n\n### Patches\n\nWe have patched the issue in eccb7ec454e6617738554a255d77f08e60ee0808 and will release TensorFlow 2.4.0 containing the patch. TensorFlow nightly packages after this commit will also have the issue resolved.\n\n### For more information\nPlease 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.\n\n### Attribution\nThis vulnerability has been reported in #42105\n\n## Affected packages\n\n- `tensorflow < 2.4.0`\n- `tensorflow-cpu < 2.4.0`\n- `tensorflow-gpu < 2.4.0`\n\n## Remediation\n\nUpgrade to a patched release:\n\n- `tensorflow 2.4.0`\n- `tensorflow-cpu 2.4.0`\n- `tensorflow-gpu 2.4.0`","depth":"sunlit","depthScore":33,"depthScoreParts":{"impact":32.5,"likelihood":0.2,"exploitation":0,"ransomware":0},"changes":[]}