Description d’origine
Issue summary: The AES-SIV cipher implementation contains a bug that causes it to ignore empty associated data entries which are unauthenticated as a consequence. Impact summary: Applications that use the AES-SIV algorithm and want to authenticate empty data entries as associated data can be misled by removing, adding or reordering such empty entries as these are ignored by the OpenSSL implementation. We are currently unaware of any such applications. The AES-SIV algorithm allows for authentication of multiple associated data entries along with the encryption. To authenticate empty data the application has to call EVP_EncryptUpdate() (or EVP_CipherUpdate()) with NULL pointer as the output buffer and 0 as the input buffer length. The AES-SIV implementation in OpenSSL just returns success for such a call instead of performing the associated data authentication operation. The empty data thus will not be authenticated. As this issue does not affect non-empty associated data authentication and we expect it to be rare for an application to use empty associated data entries this is qualified as Low severity issue.
Logiciels concernés
- OpenSSL ≥ 3.1.0 ≤ 3.1.1≥ 3.0.0 ≤ 3.0.9 correspondance exacte
Détails
- Publication
- 2023-07-14
- Dernière modification
- 2026-06-17
- Vecteur CVSS
CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:L/A:N- CWE
- CWE-354, CWE-287
- EPSS (probabilité d’exploitation à 30 jours)
- 0.62% (percentile 50)
Références
- cve.org/CVE-2023-2975
- nvd.nist.gov/vuln/detail/CVE-2023-2975
- git.openssl.org/gitweb/?p=openssl.git;a=commitdiff;h=00e2f5eea29994d19293ec4e8c8775ba73678...
- git.openssl.org/gitweb/?p=openssl.git;a=commitdiff;h=6a83f0c958811f07e0d11dfc6b5a6a98edfd5...
- www.openssl.org/news/secadv/20230714.txt
- www.openwall.com/lists/oss-security/2023/07/15/1
- www.openwall.com/lists/oss-security/2023/07/19/5
- security.gentoo.org/glsa/202402-08
- security.netapp.com/advisory/ntap-20230725-0004/
Suivez ces versions sur vos propres serveurs
InfraGuard collecte les versions logicielles installées sur chaque serveur et signale celles qui présentent des vulnérabilités connues. Auto-hébergé : rien ne quitte votre réseau.
Ce produit utilise des données de l’API du NVD, mais n’est ni approuvé ni certifié par le NVD. CVE® est une marque déposée de The MITRE Corporation ; les enregistrements CVE sont utilisés conformément aux conditions d’utilisation du CVE Program.