[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: Using Signature Policy in RFC-5126



There are a couple of notes regarding archive timestamp that could help you.

 

Note 4 says: “The hash is calculated over the concatenated data elements as received/stored, including the Type and Length encoding”.

 

Julien mentioned a point considered in Note 3: “Unless DER is used throughout, it is recommended that the binary encoding of the ASN.1 structures being time-stamped be preserved when being archived to ensure that the recalculation of the data hash is consistent”.

 

So. To simplify your implementation you should use DER encoding everywhere or just always keep original binary encoding. After that you have to concatenate binary representations of all the fields listed not trying to throw away Type and Length tags, or getting all certificates or crls from Certificates or crls elements of SignedData and concatenating them. Just concatenate full binary encodings of the listed elements in the mentioned order. But you have to turn over a SignerInfo sequence: you should not include outer SignerInfo Type and Length tags in hashing, but rather include every element of SignerInfo starting from version and finising with unsignedAttrs.

 

Said above is enough to add an archive timestamp attribute. Concerning verifying of this timestamp you should do the same concatenation, but throw away all archive timestamps with time later then in the timestamp being verified from unsignedAttrs element and encode it over again.

 

Pavel Smirnov

Crypto-Pro
Tel./Fax: +7 495 780-4820
WWW:
http://www.CryptoPro.ru
e-mail:
spv@xxxxxxxxxxxx

 

From: Yasir Khan [mailto:yasir.khan@xxxxxxxxxxxx]
Sent: Thursday, July 03, 2008 8:07 PM
To: Смирнов Павел Владимирович; ietf-smime@xxxxxxx
Subject: RE: Using Signature Policy in RFC-5126

 

RFC 5126 explains how to create archived time stamp i.e.

 

 

ArchiveTimeStampToken ::= TimeStampToken

 

   The value of the messageImprint field within TimeStampToken shall be

   a hash of the concatenation of:

 

      - the encapContentInfo element of the SignedData sequence;

 

      - any external content being protected by the signature, if the

        eContent element of the encapContentInfo is omitted;

 

      - the Certificates and crls elements of the SignedData sequence,

        when present, and;

 

      - all data elements in the SignerInfo sequence including all

        signed and unsigned attributes.

 

 

The ArchiveTimeStamp will be added as an unsigned attribute in the SignerInfo sequence.

 

According to the RFC, it is clear that message imprint will be the hash of above concatenated values but it is not explained how to concatenate these values? Please clarify this point. Some hint for implementing this would be highly appreciated. Thanks!

 

----------------------------------------------------------------------------------

YASIR KHAN
Development Manager

 

Ascertia Limited
www.ascertia.com  

www.globaltrustfinder.com

----------------------------------------------------------------------------------

Delivering Trust for e-business Documents & Workflows

----------------------------------------------------------------------------------

NOTICE: This message is intended for the recipient(s) only. This communication may contain privileged or other

confidential information. If you are not the intended recipient, or believe that you have received this communication in

error, please do not print, copy, retransmit, disseminate, or otherwise use the information. Please notify the sender that

you have received this email in error, and delete the copy you received

 


From: owner-ietf-smime@xxxxxxxxxxxx [mailto:owner-ietf-smime@xxxxxxxxxxxx] On Behalf Of Pavel V. Smirnov
Sent: Thursday, June 26, 2008 1:04 PM
To: 'Yasir Khan'; ietf-smime@xxxxxxx
Subject: RE: Using Signature Policy in RFC-5126

 

Yes, I mean “digital signature policy” as a “policy document” present at SPUri. But I use SPUri only for example, you may specify a SigPolicyId and omit sigPolicyQualifiers at all. In this case you have to rely on some other means to convey a correspondence between SigPolicyIds and “policy documents”.

 

In case of SPuri you have to retrieve a “policy document” by its URI and hash it.

 

Note that if your infrastructure provides a trusted source of “policy documents”, and there will never be two different (versions of) “policy documents” identified by the same OID, you don’t have to hash them at all. Just use zero policy hash value.

 

Pavel Smirnov

Crypto-Pro
Tel./Fax: +7 495 780-4820
WWW:
http://www.CryptoPro.ru
e-mail:
spv@xxxxxxxxxxxx

 

From: owner-ietf-smime@xxxxxxxxxxxx [mailto:owner-ietf-smime@xxxxxxxxxxxx] On Behalf Of Yasir Khan
Sent: Thursday, June 26, 2008 12:19 PM
To: Смирнов Павел Владимирович; ietf-smime@xxxxxxx
Subject: RE: Using Signature Policy in RFC-5126

 

You wrote: ”You have to hash a digital signature policy represented as a sequence of bytes in some format and place the computed value in SigPolicyHash.”

 

To which item you are naming as “digital signature policy”. You mean policy document present at SPUri? If Yes then it makes some sense. But if only SPUserNotice is present or nothing is present as sigPolicyQualifiers as it is an OPTIONAL element:

 

            sigPolicyQualifiers SEQUENCE SIZE (1..MAX) OF SigPolicyQualifierInfo OPTIONAL

 

then on which item the hash would be computed?

 

Regards,

Yasir Khan

 

 


From: Pavel V. Smirnov [mailto:spv@xxxxxxxxxxxx]
Sent: Wednesday, June 25, 2008 4:07 PM
To: 'Yasir Khan'; ietf-smime@xxxxxxx
Subject: RE: Using Signature Policy in RFC-5126

 

Hello Yasir,

 

There is no need to protect by SigPolicyHash other fields of SignaturePolicyId structure because it is placed in a signed attribute. All signed attributes are protected by the signature itself.

 

In most cases the policy would be an external document not included in your signed message, and you have to unambiguously indicate specific policy with respect to which your document should be treated. E.g., you may only have an URI pointing to the policy as a SigPolicyQualifier.

 

You have to hash a digital signature policy represented as a sequence of bytes in some format and place the computed value in SigPolicyHash.

 

Pavel Smirnov

Crypto-Pro
Tel./Fax: +7 495 780-4820
WWW:
http://www.CryptoPro.ru
e-mail:
spv@xxxxxxxxxxxx

 

From: owner-ietf-smime@xxxxxxxxxxxx [mailto:owner-ietf-smime@xxxxxxxxxxxx] On Behalf Of Yasir Khan
Sent: Wednesday, June 25, 2008 2:43 PM
To: ietf-smime@xxxxxxx
Subject: Using Signature Policy in RFC-5126

 

We have a question related to using the signature policy in the CAdES signatures (EPES) defined in RFC-5126. Here is the relevant structure:

 

SignaturePolicyId ::= SEQUENCE {

            sigPolicyIdentifier SigPolicyId,

            sigPolicyHash SigPolicyHash,

            sigPolicyQualifiers SEQUENCE SIZE (1..MAX) OF

            SigPolicyQualifierInfo OPTIONAL

}

 

SigPolicyId ::= OBJECT IDENTIFIER

 

SigPolicyHash ::= OtherHashAlgAndValue

 

OtherHashAlgAndValue ::= SEQUENCE {

            hashAlgorithm   AlgorithmIdentifier,

        hashValue       OtherHashValue

}

 

SigPolicyQualifierInfo ::= SEQUENCE {

            sigPolicyQualifierId SigPolicyQualifierId,

            sigQualifier ANY DEFINED BY sigPolicyQualifierId

}

 

SigPolicyQualifierId ::= OBJECT IDENTIFIER

id-spq-ets-uri OBJECT IDENTIFIER ::= {

            iso(1) member-body(2) us(840) rsadsi(113549) pkcs(1) pkcs9(9) smime(16) id-spq(5) 1

}

SPuri ::= IA5String

 

id-spq-ets-unotice OBJECT IDENTIFIER ::= {

            iso(1) member-body(2) us(840) rsadsi(113549) pkcs(1) pkcs9(9) smime(16) id-spq(5) 2

}

SPUserNotice ::= SEQUENCE {

            noticeRef NoticeReference OPTIONAL,

            explicitText DisplayText OPTIONAL

}

 

NoticeReference ::= SEQUENCE {

            organization DisplayText,

            noticeNumbers SEQUENCE OF INTEGER

}

 

DisplayText ::= CHOICE {

            visibleString VisibleString (SIZE (1..200)),

            bmpString BMPString (SIZE (1..200)),

            utf8String UTF8String (SIZE (1..200))

}

 

In the given structure for CAdES-EPES signature, its is not clear that whether are we computing the hash "SigPolicyHash" over the document at "SPuri" and/or over the "SPUserNotice"

 

Are the following combinations valid?

 

1) Only compute hash over document present at SPURI if only SPUri is set

2) Only compute hash over SPUserNotice  if only SPUserNotice is set

3) Compute hash over document at SPURI and SPUserNotice if both are set

 

Please clarify it. Thanks!

 

Regards,

 

Yasir Khan
Development Manager

 

Ascertia Ltd
40 Occam Road
Surrey Research Park
Guildford
Surrey, GU2 7YG
United Kingdom

 

t.  +44 (0)1483 685500
f.  +44 (0)1483 573704

 

www.ascertia.com   

 

-----------------------------------------------------------------
         Identity Proven, Trust Delivered
-----------------------------------------------------------------