Hi Wei,
> which multiple-inherits from SymmetricCipher and
> MessageAuthenticationCode.
My first stab at CCM using inheritance ("IS-A") failed miserably. I
punted and used composition ("HAS-A").
> The Filter wrapper for AuthenticatedSymmetricCipher will
> probably take confidential data ...
So far, my use of EtA has been with authenticating the confidential
data. I have not had a need to authenticate the associated data (or
AAD in GCM-speak), so a NULL pointer is fed to the formatting
function. But a proper implementation will be very useful for the
future.
Final question: I suspect the modes will not be FIPS compliant due to
the lack of validation (unless of course it is submitted for CMVP). My
thinking is that the primitives, such as AES, are FIPS compliant.
However, the additional code to thwart timing attacks on a failed
decryption-verfication will not have been certified. Can you confirm?
Thanks for rolling the implementations,
Jeff
On 2/5/09, Wei Dai <[email protected]> wrote:
> Hi Jeff. I'm already working on implementing GCM and CCM, which will
> implement AuthenticatedSymmetricCipher, which multiple-inherits from
> SymmetricCipher and MessageAuthenticationCode. The Filter wrapper for
> AuthenticatedSymmetricCipher will probably take confidential data on the
> main channel, and take additional authenticated data on a channel named
> "AAD".
>
> Please let me know if you have any other suggestions.
>
> --------------------------------------------------
> From: "Jeffrey Walton" <[email protected]>
> Sent: Wednesday, February 04, 2009 3:35 PM
> To: "Crypto++" <[email protected]>
> Subject: Question on Pipelining (Multpile Put() on discrete data elements)
>
> >
> > Hi All/Wei,
> >
> > When attemtping to pipeline in Crypto++, how do we differentiate
> > between between discrete data elements? Most filters are aware of a
> > single data element. But how do we handle the case of GCM or CCM where
> > we must distinguish between the authenticated data and the
> > confidential data?
> >
> > I suspect the use of Channels may be part of the solution, but nothing
> > is jumping out at me.
> >
> > Jeff
> >
--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the "Crypto++ Users"
Google Group.
To unsubscribe, send an email to [email protected].
More information about Crypto++ and this group is available at
http://www.cryptopp.com.
-~----------~----~----~----~------~----~------~--~---
