Re: nss RSA forgery (CVE-2014-1568)

[Hanno Böck <hanno () hboeck de>]

On Thu, 25 Sep 2014 23:17:58 +0200
Hanno Böck <hanno () hboeck de> wrote:

> Some more info, but not much:
> http://www.intelsecurity.com/advanced-threat-research/
>
> Initially I thought this is probably not such a big deal because the
> bleichenbacher attack requires e to be very small (like e=3 or e=4)
> and everyone uses e=65537 these days. But it seems I was wrong on
> that, lots of CAs still with e=3.

Correcing me: There are not lots of, but exactly 6.

Camerfirma_Chambers_of_Commerce_Root.pem: 3 (0x3)
Camerfirma_Global_Chambersign_Root.pem: 3 (0x3)
Digital_Signature_Trust_Co._Global_CA_1.pem: 3 (0x3)
Digital_Signature_Trust_Co._Global_CA_3.pem: 3 (0x3)
Go_Daddy_Class_2_CA.pem: 3 (0x3)
Starfield_Class_2_CA.pem: 3 (0x3)


Some background: having very small exponents for RSA isn't a problem if
you do everything right, but it vastly increases the likelyhood of
issues like this popping up. It'd seem like a good safeguard to get rid
of all e=3 keys.

-- 
Hanno Böck
http://hboeck.de/

mail/jabber: hanno () hboeck de
GPG: BBB51E42

Attachment: signature.asc
Description: