RE: More examples-05 issues

["Pawling, John" <John.Pawling@xxxxxxxxxxxxxxxx>]

Alexei,

I am glad to here that the updated 6.2.bin file worked as expected.  

We agree that RFC 2630 does not explicitly define which key wrap algorithm
should be used in mixed cases such as when an RC2 CEK is wrapped with 3DES
KEK.  We also agree that the key wrap algorithm must be selected based on
the CEK algorithm (i.e. RC2 key wrap must be used with RC2 CEK and 3DES key
wrap must be used with 3DES CEK).  We will use the S/MIME Freeware Library
(SFL) to generate a new section 6.9 sample message using the RC2 key wrap
algorithm to wrap the RC2 CEK.

The SFL allows the calling application to select the KEK algorithm and the
key wrap algorithm separately from the content encryption algorithm.  We
will enhance the SFL to force compliance that the key wrap algorithm must be
selected based on the CEK algorithm.

Thank you again for your feedback!!
===========================================
John Pawling, John.Pawling@xxxxxxxxxxxxxxxx
Getronics Government Solutions, LLC
===========================================

 

-----Original Message-----
From: Alexei Shamov [mailto:shamov@xxxxxxx]
Sent: Wednesday, January 17, 2001 12:07 PM
To: 'Pawling, John'; ietf-smime-examples@xxxxxxx
Cc: 'Colestock, Robert'
Subject: RE: More examples-05 issues


Hi John,

Thank you for the update. Please find my comments below.

<Snip>

>[JSP:  Section 6.2 contains a sample EnvelopedData using TripleDES for
>encrypting and RSA for key management, so I do not understand your above
>comment in relation to the section 6.2 sample.   In the section 6.9 sample
>message, a 3DES KEK is used, so the Triple-DES key wrap algorithm is used.]

RFC 2630 does not explicitly define which algorithm should be used in mixed
cases, when RC2 CEK is wrapped with 3DES KEK.

However, 3DES key wrap can not be used in 6.9 because of the following:
1. 3DES unwrap algorithm MUST fail at 12.6.3.1 simply because ciphertext is
not 40 bytes long.
2. 3DES key parity adjustment/verification does not make sence for RC2 CEK.

I think that wrap algorithm should be selected according to CEK algoritm,
not KEK algorithm (ie RC2 key wrap (12.6.4) should be used with RC2 CEK,
and 3DES key wrap (12.6.2) with 3DES CEK).

Btw. in the KEKRecipientInfo of the same message rc2 key wrap was correcly
selected:

01E1 A2   63:         [2] {
01E3 02    1:           INTEGER 4
01E6 30   22:           SEQUENCE {
01E8 04   11:             OCTET STRING 'MailListTripleDES'
01FB 18    D:             GeneralizedTime '951230235959Z'
            :             }
020A 30   10:           SEQUENCE {
020C 06    B:             OBJECT IDENTIFIER
            :               id-alg-CMSRC2wrap (1 2 840 113549 1 9 16 3 7)
0219 02    1:             INTEGER 58
            :             }

Regards,
Alexei