On Mon, Apr 10, 2023 at 10:40 AM Ben Beasley code@musicinmybrain.net wrote:
When I took over maintenance of the flintqs package[1]—which contains William Hart’s quadratic sieve implementation, as modified for sagemath—I built it for EPEL7, EPEL8, and EPEL9. My thoughts were, “Why not? Someone might find it useful.”
It was recently pointed out[2][3] that the flintqs command-line tool uses temporary files in unsafe ways[4], which could potentially represent an exploitable security vulnerability; this has been assigned CVE-2023-29465[5].
There is no immediate patch available; while one could surely be constructed, the sagemath project plans to incorporate the factorization algorithm directly in sagemath and discontinue support of the vulnerable command-line tool rather than fixing it[6].
Since sagemath is not packaged in any of the EPEL releases, and flintqs is therefore a leaf package, I plan to handle this security report by retiring flintqs in all three EPELs. This email is the beginning of that process as prescribed in the EPEL Retirement Policy: Process: Security Reasons[7]. I doubt there will be any objections, but the process requires a one-week discussion period, so I will follow up on the epel-announce list and do the retirements no earlier than 2023-03-17.
[1] https://src.fedoraproject.org/rpms/flintqs
[2] https://bugzilla.redhat.com/show_bug.cgi?id=2185301
[3] https://github.com/sagemath/FlintQS/issues/3
[4] https://owasp.org/www-community/vulnerabilities/Insecure_Temporary_File
[5] https://nvd.nist.gov/vuln/detail/CVE-2023-29465
[6] https://github.com/sagemath/sage/pull/35419
[7]
https://docs.fedoraproject.org/en-US/epel/epel-policy-retirement/#process_se...
Thank you for following the retirement policy.
I'm assuming that's a typo and you really meant "no earlier than 2023-04-17"
Troy