CVE-2020-28975
First published: Sat Nov 21 2020(Updated: )
** DISPUTED ** svm_predict_values in svm.cpp in Libsvm v324, as used in scikit-learn 0.23.2 and other products, allows attackers to cause a denial of service (segmentation fault) via a crafted model SVM (introduced via pickle, json, or any other model permanence standard) with a large value in the _n_support array. NOTE: the scikit-learn vendor's position is that the behavior can only occur if the library's API is violated by an application that changes a private attribute.
Credit: cve@mitre.org cve@mitre.org
| Affected Software | Affected Version | How to fix |
|---|---|---|
| pip/scikit-learn | >=0.23.2<1.0.1 | 1.0.1 |
| Scikit-learn | >=0.23.2<1.0.1 |
Never miss a vulnerability like this again
Sign up to SecAlerts for real-time vulnerability data matched to your software, aggregated from hundreds of sources.
Reference Links
- http://packetstormsecurity.com/files/160281/SciKit-Learn-0.23.2-Denial-Of-Service.html
- http://seclists.org/fulldisclosure/2020/Nov/44
- https://github.com/cjlin1/libsvm/blob/9a3a9708926dec87d382c43b203f2ca19c2d56a0/svm.cpp#L2501
- https://github.com/scikit-learn/scikit-learn/commit/1bf13d567d3cd74854aa8343fd25b61dd768bb85
- https://github.com/scikit-learn/scikit-learn/issues/18891
- https://security.gentoo.org/glsa/202301-03
- https://nvd.nist.gov/vuln/detail/CVE-2020-28975
- https://github.com/advisories/GHSA-jxfp-4rvq-9h9m (Advisory)
- https://github.com/pypa/advisory-database/tree/main/vulns/scikit-learn/PYSEC-2020-108.yaml
Frequently Asked Questions
What is the severity of CVE-2020-28975?
CVE-2020-28975 has been reported as a denial of service vulnerability that can lead to a segmentation fault.
How do I fix CVE-2020-28975?
To fix CVE-2020-28975, upgrade scikit-learn to version 1.0.1 or later.
Which software is affected by CVE-2020-28975?
CVE-2020-28975 affects scikit-learn versions between 0.23.2 and 1.0.1.
What does CVE-2020-28975 impact?
CVE-2020-28975 impacts the svm_predict_values function in the Libsvm implementation.
Can CVE-2020-28975 be exploited remotely?
Yes, CVE-2020-28975 can be exploited remotely through crafted model inputs.
- agent/type
- agent/severity
- agent/remedy
- agent/status
- agent/author
- agent/softwarecombine
- agent/description
- agent/first-publish-date
- agent/event
- collector/mitre-cve
- source/MITRE
- collector/github-advisory-latest
- source/GitHub
- alias/GHSA-jxfp-4rvq-9h9m
- alias/CVE-2020-28975
- agent/software-canonical-lookup
- agent/last-modified-date
- agent/references
- collector/github-advisory
- agent/trending
- agent/source
- agent/tags
- collector/nvd-api
- source/NVD
- agent/software-canonical-lookup-request
- collector/nvd-cve
- collector/nvd-index
- package-manager/pip
- vendor/scikit-learn
- canonical/scikit-learn
- version/scikit-learn/0.23.2
- status/disputed