Hide Forgot
http://www.mozilla.org/projects/security/pki/nss/nss-3.11/nss-3.11-algorithms.html notes that NSS (or its soft token, at least) limits DH key agreement to situations where the modulus is at most 2236 bits. This is still the value of DH_MAX_P_BITS in nss-3.12.8-1.el5. The OpenSSL-based PKINIT implementation in MIT Kerberos can be configured to attempt to do so with a modulus of up to 4096 bits in size (Oakley group 16), so this is a potential interoperability problem for NSS-based PKINIT implementations. Is it possible to raise this limit?
The max size of the keys was increased to 16384 but it wan't until NSS-3.14. https://bugzilla.mozilla.org/show_bug.cgi?id=596692 https://bugzilla.mozilla.org/show_bug.cgi?id=636802. On RHEL-5.x because of the need to preserve the FIPS validation softoken/freebl is at 3.11.5. Even RHEL-6.x is at softoken from 3.12.9, for FIPS reasons also, so it misses out on this change as well.
This request was evaluated by Red Hat Product Management for inclusion in a Red Hat Enterprise Linux release. Product Management has requested further review of this request by Red Hat Engineering, for potential inclusion in a Red Hat Enterprise Linux release for currently deployed products. This request is not yet committed for inclusion in a release.
Could you please set devel_ack/condnack for 5.10 or postpone/close this bugzilla? Erratas should be filed by May 21st, so development phase is nearing its exit.
(In reply to Ondrej Vasik from comment #7) > Could you please set devel_ack/condnack for 5.10 or postpone/close this > bugzilla? Erratas should be filed by May 21st, so development phase is > nearing its exit. (In reply to Karel Srot from comment #8) > Hi, > may I ask for a devel review with respect to 5.10 inclusion? This bug needs to be included. In fact, the needed changes are actually committed already as consequence of having completely rebased nss to 3.14.3 to address Lucky 13.
Elio: If it needs to be included, please get devel_ack for it.
r+ rrelyea, this happened as part of the rebase anyway, but as well clear the bug report out. bob
Per comment #c21 this should be already fixed in RHEL-5.10 wit the rebase. Please reopen the bug if you find that not being correct.