CVE-2020-15193 json
Memory corruption in Tensorflow
published 2020-09-25 6 y ago modified 2026-09-08 2 mo ago also BIT-tensorflow-2020-15193 , EUVD-2020-0189 , GHSA-rjjg-hgv6-h69v , PYSEC-2020-116 , PYSEC-2020-273 , PYSEC-2020-308
- cvss
- 7.1
- version 3.1, CNA
- cisa kev
- Not listed
- not in CISA's catalog
- epss
- 0.8%
- EPSS 0.8% chance of exploitation in 30 days, percentile 57%
- public exploits
- 0
- none known
Summary
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.
Patch / fix
- pypi tensorflow 2.2.1, 2.3.1
- pypi tensorflow-cpu 2.2.1, 2.3.1
- pypi tensorflow-gpu 2.2.1, 2.3.1
- bitnami tensorflow 2.2.1, 2.3.1
- tensorflow2 openSUSE Leap 15.2 2.1.2-lp152.7.3.1
- tensorflow2-lite openSUSE Leap 15.2 2.1.2-lp152.7.3.1
- tensorflow2_2_1_2-gnu-hpc openSUSE Leap 15.2 2.1.2-lp152.7.3.1
- tensorflow2_2_1_2-gnu-openmpi2-hpc openSUSE Leap 15.2 2.1.2-lp152.7.3.1
opensuse advisories openSUSE-SU-2020:1766-1
Exploit prediction
0.83% chance of exploitation in the next 30 days, higher than 57% of all scored CVEs (FIRST EPSS, 2026-11-08).
Exploits and checks
- github.com https://github.com/tensorflow/tensorflow/security/advisories/GHSA-rjjg-hgv6-h69v referenced PoC, tagged "exploit" by a source, unverified
CVSS scores
| assigned by | version | score | vector |
|---|---|---|---|
| CNA | 3.1 | 7.1 | CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:N/I:H/A:L |
| GitHub | 3.1 | 7.1 | CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:N/I:H/A:L |
| NVD | 3.1 | 7.1 | CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:N/I:H/A:L |
| NVD | 2.0 | 5.5 | AV:N/AC:L/Au:S/C:N/I:P/A:P |
| OSV | 3.1 | 7.1 | CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:N/I:H/A:L |
Affected packages
| ecosystem | package | affected | fixed in |
|---|---|---|---|
| pypi | tensorflow | >=2.2.0, <2.2.1 >=2.3.0, <2.3.1 <2.2.1 >=2.3.0rc0, <2.3.1 |
2.2.1 2.3.1 |
| pypi | tensorflow-cpu | >=2.2.0, <2.2.1 >=2.3.0, <2.3.1 <2.2.1 >=2.3.0rc0, <2.3.1 |
2.2.1 2.3.1 |
| pypi | tensorflow-gpu | >=2.2.0, <2.2.1 >=2.3.0, <2.3.1 <2.2.1 >=2.3.0rc0, <2.3.1 |
2.2.1 2.3.1 |
| bitnami | tensorflow | >=2.2.0, <2.2.1 >=2.3.0, <2.3.1 |
2.2.1 2.3.1 |
Affected products
- tensorflow tensorflow
Classification
- Weakness
-
- CWE-908 Use of Uninitialized Resource
References
-
github.com
/tensorflow/tensorflow/commit/22e07fb204386768e5bcbea563641ea11f96ceb8
web, patch, third-party-advisory, advisory
-
github.com
/tensorflow/tensorflow/security/advisories/GHSA-rjjg-hgv6-h69v
web, exploit, third-party-advisory, advisory
-
lists.opensuse.org
/opensuse-security-announce/2020-10/msg00065.html
vendor-advisory, web, mailing-list, third-party-advisory, advisory
-
github.com
/tensorflow/tensorflow/releases/tag/v2.3.1
web, third-party-advisory, advisory
-
nvd.nist.gov
/vuln/detail/CVE-2020-15193
advisory, web
-
github.com
/pypa/advisory-database/tree/main/vulns/tensorflow-cpu/PYSEC-2020-273.yaml
web
-
github.com
/pypa/advisory-database/tree/main/vulns/tensorflow-gpu/PYSEC-2020-308.yaml
web
-
github.com
/pypa/advisory-database/tree/main/vulns/tensorflow/PYSEC-2020-116.yaml
web
-
github.com
/tensorflow/tensorflow
package
Sources
- CVE List CVE-2020-15193
- GitHub Advisory GHSA-rjjg-hgv6-h69v
- NVD CVE-2020-15193
- OSV BIT-tensorflow-2020-15193
- OSV PYSEC-2020-116
- OSV PYSEC-2020-273
- OSV PYSEC-2020-308
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.