← Prev in month ← Prev in thread
Next in thread → Next in month →

PKCS11 base spec 3.0 WD 05 comments.

From
Robert Relyea <>
Date
2018-07-31T00:25:59+00:00
ID
Thread
PKCS11 base spec 3.0 WD 05 comments.
1. It looks like my feedback from 1/24 didn't quite get
      incorporated:
    

      My comments are basically deficiencies
        with the original proposal (which becomes more clear with the
        proposals being added to the spec.
        

        

        Pages 37 and 38.
        

        

         - C_GetFunctionLists and it's description (starting with
        "C_GetFunctionLists(" and ending with "only an example" on page
        82 should be added in the function list section of the base
        spec, right after C_GetFunctionList on page 82.
        

        

        ÂÂ - Starting at '"PKCS 11 3.0" returns a new function list with
        is the same as... ' until 'This will be defined as
        CK_FUNCTION_LIST_2_42' replace with '"PKCS 11 3.0 returns a
        CK_FUNCTION_LIST_3_0".
        

        

        Â - The section described above was supposed to be instructions
        to the editor for defining CK_FUNCTION_LIST_3_0, which should go
        here (right after CK_FUNCTION_LISTS definition. Also
        C_SessionCancel and C_LoginUser should be added to the new
        function table list.
        

        

        Page 66
        

        

        - Table 30. Add C_GetFunctionLists in "General Purpose
        Functions" right after C_GetFunctionList. The description should
        be "Obtain various function entry points of the Cryptoki library
        identified by interface string" (I'm OK with serious editing of
        the previous description by the editor).
        

      
      Here's another attempt to describe the issue:

    
The "New function proposal" got incorporated as a full cut and
      past into the document. I basically has 3 separate pieces:

    
1) The new structures (CK_INTERFACES and CK_FUNCTION_LISTS).
      Those have been incorporated in the correct point in the document,
      no changes there.

    
2) The descripton of the C_GetFunctionLists function. This runs
      from "C_GetFunctionLists(" to the end of the spec (ending at "...
      is only an example."), Excluding the <NOTE:....>. Moving
      this to just behind the C_GetFunctionList and formatting properly
      should be sufficient. There are bugs in this text which I will
      point out in additional bullets.

    

    
3) <NOTE:...> is a note that section 2.5.2 of the pkcs11
      User Guide (pkc11-ug) needs to be updated to point to the change
      in semanatics id the CKF_FORK_SAFE_INTERFACE is defined. For this
      spec it can be removed.

    

    
Also the proposal implies:

    
Âa. a new function table type: CK_FUNCTION_LIST_3_0 (there's a
      bug in the proposal and it calls it CK_FUNCTION_LIST_2_42, but
      that was a typo). That list should have all the functions in
      CK_FUNCTION_LIST, plus all the new functions we added in 3.0. The
      new functions should be at the end of the list. Those function are
      C_GetFunctionLists, C_LoginUser,
      C_MessageEncryptInit().....C_MessageVerifyFinal().

    

    
b. C_GetFunctionLists need to be added to the list of function in
      table 30.

    2. CK_FUNCTION_LIST should not have the new v3.0 functions in
      it. The new functions should be added to a new
      CK_FUNCTION_LIST_3_0, and they should be added at the end.
    
This is a C struct, so order matters from a compatibility point
      of view.

    3. The original "New function proposal" is had some errors which
      were incorporated in this spec:
    
a. CK_FUNCTION_LIST_2_42 should be CK_FUNCTION_LIST_3_0.

    
b. C_LoginUser() should be added to the list of function between
      C_GetFunctionLists and C_MessageEncryptInit().

    

    4. Typo:CKO_PRIFILE should be CKO_PROFILE in section 4.13.2
    5. The heading for C_MessageEncryptFile has Message and Encrypt
      transposed.
    6. C_MessageSign* and C_MessageVerify* are missing from the body
      of the spec.
    7. Source Spec AEAD message was an older draft and missing an
      update the the C_*Next functions, namely adding a parameter
      pointers to the Next functions. The draft with these changes can
      be found here:Â
https://www.oasis-open.org/apps/org/workgroup/pkcs11/download.php/60200/AEAD_proposal.docÂ
      The changes are marked in the document with change bars, so it
      should be relatively simple to pick up.
← Prev in month ← Prev in thread
Next in thread → Next in month →