Jim:

I was thinking of the first case, but it would be good to support both.

Russ


> On Jan 22, 2019, at 6:33 PM, Jim Schaad <[email protected]> wrote:
> 
> Jim,
> 
> Ok lets try this again, I don't think that you got the gist of my question.
> (I have been told recently that I probably need to be more verbose in
> describing my questions.)
> 
> What I understand you want to do is to build a structure that looks
> something like this:
> 
> EncryptedMessage : {
>   Protected Headers,
>  Unprotected Headers,
>  AES-KeyWrap-Pad( PKCS #8 ),
>  Recipient: [ {
>       Protected Headers,
>      Unprotected Headers,
>      RSA-OAEP( AES-KeyWrap-Pad encryption key )
>  }]
> }
> 
> In this case the key wrap is being done as a content encryption key even if
> what it is encrypting a content which contains a private key.
> 
> 
> The other case that I was talking about would be something like the
> following:
> 
> EncryptedMessage :{
>    Protected Headers,
>    Unprotected Headers,
>    AES-GCM-128( some content ),
>    Recipient: [ {
>          Protected Headers,
>          Unprotected Headers,
>           ECDH ( CEK ),
>             Recipient: [{
>                 Protected Headers,
>                 Unprotected Headers,
>                 AES-KeyWrap-Pad( PKCS#8 w/ private ECDH key )
> }]}]}
> 
> In this case one would both be transporting the private key - encrypted with
> AES-KeyWrap and a pre-shared secret (or perhaps ECDH and AES-KeyWrap-Pad)
> and that transported key is then being used to do an ECDH operation which is
> then used to decrypt some content.
> 
> In this case one is in essence transporting two different sets of encrypted
> material:  the "some content" information and the private key.
> 
> Jim
> 
> 
>> -----Original Message-----
>> From: Russ Housley <[email protected]>
>> Sent: Tuesday, January 22, 2019 3:21 PM
>> To: Jim Schaad <[email protected]>
>> Cc: cose <[email protected]>
>> Subject: Re: [COSE] Adding AES Key Wrap with Padding
>> 
>> Jim:
>> 
>> PKCS#8 is a key that need to be wrapped.
>> 
>> I want to be able to use if for any keying material, including an HMAC
> key.
>> 
>> Russ
>> 
>> 
>>> On Jan 22, 2019, at 6:10 PM, Jim Schaad <[email protected]> wrote:
>>> 
>>> Russ,
>>> 
>>> That is indeed an interesting use case.  Does this mean that you are
>>> really more interested in using this as a content encryption algorithm
>>> rather than a key encryption algorithm?  I have doubts for this use
>>> case that you are looking to wrap key and are looking to wrap content.
>>> This is different than a use case where one wishes to use an HMAC
>>> algorithm with a key which is not a multiple of 8 octets.
>>> 
>>> Jim
>>> 
>>> 
>>>> -----Original Message-----
>>>> From: Russ Housley <[email protected]>
>>>> Sent: Tuesday, January 22, 2019 2:36 PM
>>>> To: Jim Schaad <[email protected]>
>>>> Cc: cose <[email protected]>
>>>> Subject: Re: [COSE] Adding AES Key Wrap with Padding
>>>> 
>>>> Jim:
>>>> 
>>>> The use case is wrapping PKCS#8 objects, which are not necessarily a
>>> multiple
>>>> of 8 octets.
>>>> 
>>>> Russ
>>>> 
>>>> 
>>>>> On Jan 22, 2019, at 5:09 PM, Jim Schaad <[email protected]>
> wrote:
>>>>> 
>>>>> Russ,
>>>>> 
>>>>> I will ask the same question that I asked the last time that
>>>>> somebody requested that we support the AES Key Wrap with Padding
>> Algorithm.
>>>>> What is the use case that you have for this algorithm.  This makes a
>>>>> difference in terms of how things are going to be used and what
>>>>> other things might need to be specified for this.  I have to admit
>>>>> that I would really prefer to get an AEAD version of this algorithm
>>>>> if we are going to add something but that I probably not possible.
>>>>> 
>>>>> Jim
>>>>> 
>>>>> 
>>>>> _______________________________________________
>>>>> COSE mailing list
>>>>> [email protected]
>>>>> https://www.ietf.org/mailman/listinfo/cose
>>> 
> 

_______________________________________________
COSE mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/cose

Reply via email to