Fixes for CKA_PUBLIC_KEY_INFO

[Michael StJohns <msj@nthpermutation.com>]

In the changes for CKA_PUBLIC_KEY_INFO there was a line that said that 
the format of the public key info would be specified under the key type 
definitions.  BUT - those weren't added.  Here are the proposed edits.  
Note there are a few other items (see ">>>>>") that require some 
consideration as well.


Changes:

RSA  - Insert this section after 2.1.3 and adjust the numbering:

"2.1.4 RSA Encoding for CKA_PUBLIC_KEY_INFO

The format for the subjectPublicKey field of the SubjectPublicKeyInfo 
structure for an RSA key is specified in section 1.2 of [RFC4055]."

>>> Is there a PKCS11 definition for an RSA-OAEP public key??

DSA - Insert section after 2.2.4 and adjust the numbering:

"2.2.5 DSA Key Encoding For CKA_PUBLIC_KEY_INFO

The format for the subjectPublicKey field of the SubjectPublicKeyInfo 
structure for a DSA key is specified in section 2.3.2 of [RFC3279].  The 
parameters field of the algorithm field of the structure SHOULD present 
and, if present, the components of that field must have values identical 
to those in the matching CKA_PRIME, CKA_SUBPRIME and CKA_BASE 
attributes. If the parameters field is not present, the parameters are 
taken from the CKA_PRIME, CKA_SUBPRIME and CKA_BASE attributes of the key.

EC insert section after 2.3.4 and adjust:

"2.3.5 EC Key Encoding  for CKA_PUBLIC_KEY_INFO

The format for the subjectPublicKey field of the SubjectPublicKeyInfo 
structure for an Elliptic Curve (EC) key is specified in sections 2.1.1 
and 2.2 of [RFC5480].  The parameters field of the algorithm field of 
the structure SHOULD be present, and if present MUST contain the same 
value as the CKA_EC_PARAMS attribute.  If the field is absent, the 
parameters for the key are taken from CKA_EC_PARAMS."

DH Insert section after 2.4.5 and adjust:

"2.4.6 DH Key Encoding for CKA_PUBLIC_KEY_INFO

The format for the subjectPublicKey field of the SubjectPublicKeyInfo 
structure for a Diffie-Hellman key or Fiffie-Hellman X9.42 key is 
specified in section 2.3.3. of [RFC3279]."


>>>>>>>  One of the required fields for the above encoding is the Q or 
CKA_SUBPRIME value which isn't present in the plain DH public key (but 
is present in the plain DH private key.  I think this should be added as 
a mandatory field similar to what we did with adding CKA_PUBLIC_EXPONENT 
to the RSA private key object.


GOST R 34.10-2001  Insert section after 2.48.3 and adjust:

"2.48.4 GOST R 34.10-2011 Format for CKA_PUBLIC_KEY_INFO

The format for the subjectPublicKey field of the SubjectPublicKeyInfo 
structure for a GOST R 34.10-2001 key is specified in section 2.3.2 of 
[RFC4491].  The parameters field of the algorithm field of the structure 
SHOULD be present and, if present, MUST contain the same value as the 
CKA_GOSTR3410PARAMS attribute of the key.  If parameters field is 
absent, the parameters for the key are taken from CKA_GOSTR3410PARAMS."