3 ms·
Is it possible/likely that many of these bad pairs are from this debian/openssl bug? http://taint.org/2008/05/13/153959a.html http://taint.org/2008/05/13/15395
by robrenaud 15y ago
Is it possible/likely that many of these bad pairs are from this debian/openssl bug?
http://taint.org/2008/05/13/153959a.html http://taint.org/2008/05/13/153959a.html
- serialx 15y agoNone of the newly found aected moduli are blacklisted (cf. [24]). Based on this brief analysis, we conclude that the problem reported in the main body of this paper persists and that the 99.8% security level of 1024-bit RSA moduli is further reduced. pp. 15 http://eprint.iacr.org/2012/064.pdf http://eprint.iacr.org/2012/064.pdf
- lambda 15y agoThey explicitly excluded the Debian/OpenSSL blacklisted keys from their sample before checking for shared factors.