Skip to content
Fencer Vulns
high 7.8

CVE-2021-41221 json

Access to invalid memory during shape inference in `Cudnn*` ops

published 2021-11-05 5 y ago modified 2026-09-08 2 mo ago also BIT-tensorflow-2021-41221 , EUVD-2021-0443 , GHSA-cqv6-3phm-hcwx , PYSEC-2021-413 , PYSEC-2021-630 , PYSEC-2021-828

cvss
7.8
version 3.1, CNA
cisa kev
Not listed
not in CISA's catalog
epss
0.2%
EPSS 0.2% chance of exploitation in 30 days, percentile 12%
public exploits
0
none known

Summary

TensorFlow is an open source platform for machine learning. In affected versions the shape inference code for the `Cudnn*` operations in TensorFlow can be tricked into accessing invalid memory, via a heap buffer overflow. This occurs because the ranks of the `input`, `input_h` and `input_c` parameters are not validated, but code assumes they have certain values. The fix will be included in TensorFlow 2.7.0. We will also cherrypick this commit on TensorFlow 2.6.1, TensorFlow 2.5.2, and TensorFlow 2.4.4, as these are also affected and still in supported range.

Patch / fix

  • pypi tensorflow 2.4.4, 2.5.2, 2.6.1, 2.7.0
  • pypi tensorflow-cpu 2.4.4, 2.5.2, 2.6.1, 2.7.0
  • pypi tensorflow-gpu 2.4.4, 2.5.2, 2.6.1, 2.7.0
  • bitnami tensorflow 2.4.4, 2.5.2, 2.6.1
  • google tensorflow 2.4.4, 2.5.2, 2.6.1
  • tensorflow Arch Linux 2.6.1-1

arch linux advisories AVG-2529

Exploit prediction

0.22% chance of exploitation in the next 30 days, higher than 12% of all scored CVEs (FIRST EPSS, 2026-11-08).

Exploits and checks

  • github.com https://github.com/tensorflow/tensorflow/security/advisories/GHSA-cqv6-3phm-hcwx referenced PoC, tagged "exploit" by a source, unverified

CVSS scores

assigned by version score vector
CNA 3.1 7.8 CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H
GitHub 3.1 7.8 CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H
NVD 2.0 4.6 AV:L/AC:L/Au:N/C:P/I:P/A:P
OSV 3.1 7.8 CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H

Affected packages

ecosystem package affected fixed in
pypi tensorflow
>=2.6.0, <2.6.1
>=2.5.0, <2.5.2
<2.4.4
>=2.7.0rc0, <2.7.0
2.6.1
2.5.2
2.4.4
2.7.0
pypi tensorflow-cpu
>=2.6.0, <2.6.1
>=2.5.0, <2.5.2
<2.4.4
>=2.7.0rc0, <2.7.0
2.6.1
2.5.2
2.4.4
2.7.0
pypi tensorflow-gpu
>=2.6.0, <2.6.1
>=2.5.0, <2.5.2
<2.4.4
>=2.7.0rc0, <2.7.0
2.6.1
2.5.2
2.4.4
2.7.0
bitnami tensorflow
>=2.4.0, <2.4.4
>=2.5.0, <2.5.2
>=2.6.0, <2.6.1
2.4.4
2.5.2
2.6.1

Affected products

  • tensorflow tensorflow

Classification

Weakness
  • CWE-120 Buffer Copy without Checking Size of Input
  • CWE-787 Out-of-bounds Write

References

Sources

Fencer Vulns keeps one entry per vulnerability: the CVE id wins, then GHSA, and records that alias each other are merged, except that two CVEs always keep separate entries. The same data is available as JSON.