Re: [pkix] I-D Action:draft-ietf-pkix-rfc3161-update-06.txt

["Andrea Caccia" <[e-mail]>]

Stefan, Nick,

in the situation where TSS is in a PKI domain where CA A is trusted, there
could (very theoretically) be a bad CA (CA B) that issue a fake certificate
to the TSS that has the same hash of the good certificate issued by CA A and
modify a TST in a way that it appears as signed by B.

Is this attack in scope? I’m not sure but possibly yes, so we can mention
it.

Is this attack feasible and is SHA-1 a problem? I think it is very difficult
to exploit such an attack because in order to obtain the fake certificate
and a “good” collision, each trial requires the computation of the fake
certificate signature to calculate the hash that must be compared with the
wanted one. The signature calculation is very time consuming and I guess
that even using a poor hash algorithm makes very difficult to exploit this
attack. 

Indeed, as a developer, this issue is not really a top priority: my concern
could be to have the possibility to choose the algorithm if this allows me
some code and test optimization. And of course in some country there could
be a policy to generally identify allowed hash algorithm that are considered
overall secure, it is more simple than specifying a different algorithm for
each use. There are IMO the real motivations to introduce ESSCerIDv2 option.

 

Andrea

 

--------
This message is sent to one or more specific recipient. If you are not the
intended recipient, please notify the sender and delete this message. 

--------
Questo messaggio è inviato a specifici destinatari. Se non siete i
destinatari, siete pregati di informare il mittente e cancellare questo
Messaggio.

 

From: [e-mail] [mailto:[e-mail]] On Behalf Of
Stefan Santesson
Sent: Tuesday, September 29, 2009 9:58 PM
To: Pope, Nick; '[e-mail]'
Cc: pkix
Subject: Re: [pkix] I-D Action:draft-ietf-pkix-rfc3161-update-06.txt

 

Nick,

BCP 72 specifies guidelines for writing security considerations sections.

It states:

  Authors MUST describe

      1.   which attacks are out of scope (and why!)
      2.   which attacks are in-scope
      2.1  and the protocol is susceptible to
      2.2  and the protocol protects against

The PKI trust model, on which we base time stamping, assumes that we are
able to determine which authorities we trust and that the data they provide
in the form of certificates are trustworthy.
As such, it is generally out of scope for a PKI based protocol to actively
defend against a trusted malicious CA. 
If you as relaying party have a problem in this area, you ought to fix how
you trust Cas, rather than trying to build protocols that is secure despite
trust in malicious CAs.

The reason is simply because the battle is lost anyway. If you trust me as
CA, and I’m a bad guy, I can screw you over in so many ways that you can’t
ever fix through protocol design.

If we are to follow BCP 72, we need to consider threats related to
ESSCertIDv2. The only one we know of is one involving a trusted malicious
CA.
Further following BCP 72 we need to decide whether the threat of a trusted
malicious CA is in scope, or out of scope.

Until we do that, and express it in writing, we are in a dead-lock position
and can’t proceed with this draft.

Agreeing on a fuzzy writing that gets the document stalled in IESG is not a
responsible path IMO. Thus I prefer a straw-poll.

/Stefan



On 09-09-29 7:18 PM, "Pope, Nick" <[e-mail]> wrote:

Stefan,
 
I don't understand the use of scope in "security considerations."  The scope
of the RFC is ESSCert2.  The discussion of security threats looks at the
wider security environment.  How can one say certain threats relating to
ESSCertv2 are out of scope?  It is reasonable that the security
considerations does not turn into a detailed analysis.
 
I do think we are so close on this area.  Given other areas of fuzziness in
RFCs I can't see oneone getting uptight about scope relating to the security
considerations.
 
Nick
 
 
 

-----Original Message-----
From: Stefan Santesson [mailto:[e-mail]] 
Sent: 29 September 2009 17:40
To: Pope, Nick; '[e-mail]'
Cc: pkix
Subject: Re: [pkix] I-D Action:draft-ietf-pkix-rfc3161-update-06.txt

Unfortunately I don't think any of them will get past the IESG review.
The obvious discuss will be "What are the threats, and are they in scope?"

/Stefan

On 09-09-29 6:24 PM, "Pope, Nick" <[e-mail]> wrote:
Either is OK for me too.
 

-----Original Message-----
From: [e-mail] [mailto:[e-mail]] 
Sent: 29 September 2009 17:19
To: Pope, Nick
Cc: '[e-mail] <[e-mail]> '; pkix
Subject: Re: [pkix] I-D Action:draft-ietf-pkix-rfc3161-update-06.txt


Nick,



I see two ways to solve and close this issue.



Solution A.



"ESSCertID provide means based on the SHA-1 hash algorithm for
identifying the certificate used to verify the signature on a time
stamp. The use of ESSCerIDv2 aims to enable implementers to comply
with policies that require phasing out all uses of the SHA-1
algorithm. Upgrading the hash function through ESSCertIDv2 address
some theoretical scenarios involving a trusted malicious CA".



Solution B.



"ESSCertID provide means based on the SHA-1 hash algorithm for
identifying the certificate used to verify the signature on a time
stamp. The use of ESSCerIDv2 aims to enable implementers to comply
with policies that require phasing out all uses of the SHA-1
algorithm. Upgrading the hash function through ESSCertIDv2 address
some theoretical scenarios involving a trusted malicious CA. This section 
does not go into a detailed analysis of the threats relating to trusted
malicious CAs".
I can live with any of them.

Denis

[e-mail] wrote: -----

To: "'[e-mail]' <[e-mail]> "
<[e-mail]>, pkix <[e-mail]>
From: "Pope, Nick" <[e-mail]>
Sent by: [e-mail]
Date: 09/29/2009 05:47PM
Subject: Re: [pkix] I-D Action:draft-ietf-pkix-rfc3161-update-06.txt

Denis,

I think the point is that this SecurityConsideration is not the place to go
into the detailed analysis of the threatsaround malicious CAs.

Perhaps a statement to this effect that "Thissection does not go into a
detailed analysis of the threats relating tomalicious CAs." could  be used
instead.

Nick

 
 

-----Original Message-----
From: [e-mail][mailto:pkix-bounces
<[e-mail]%5bmailto:pkix-bounces> @ietf.org] On Behalf Of
[e-mail]
Sent: 29 September 2009 16:42
To: pkix
Subject: Re: [pkix] I-DAction:draft-ietf-pkix-rfc3161-update-06.txt


Stefan,



You said:



" I have no problem with the dispositionchange. The problem is that .."



If you have really no problem, that let us pick myproposed wording.



However, it seems that you do have a problem, so if it isreally the case 
please do not write "I have no problem with the disposition change".



Nothing is out of scope in a security considerationssection. In particular, 
a trusted malicious CA is within the scope. The CA is trusted because
itbelongs 
to a trusted certification path, but happens to maliciously issue a
TSAcertificate.

In such a case, the original certificate can bereplaced in a log file
without anybody noticing it.



So I can't agree to have the following sentence:



"Threats caused by a trusted malicious CA ishowever out of scope of this
specification".

Denis


[e-mail]: -----

To: Dino Esposito <[e-mail]>,"Pope, Nick"
<[e-mail]>
From: Stefan Santesson <[e-mail]>
Sent by: [e-mail]
Date: 09/29/2009 02:39PM
cc: pkix <[e-mail]>, [e-mail]
Subject: Re: [pkix] I-D Action:draft-ietf-pkix-rfc3161-update-06.txt

Dino,

I agree to a certain point, but it is actually relevant in a security
considerations section to mention that a threat is out of scope.

I would also prefer the first option, but could live with both.

/Stefan

On 09-09-29 1:10 PM, "Dino Esposito"<[e-mail]> wrote:

> Nick
> I prefer the first option, so reduced
> ESSCertID provide means based on the SHA-1 hash algorithm for
> identifying the certificate used to verify the signature on a time
> stamp. The use of ESSCerIDv2 aims to enable implementers to comply
> with policies that require phasing out all uses of the SHA-1
> algorithm.
> 
> If something is out of scope, why should you address it?
> In fact, the reason to upgrade the RFC is just to enable people to
> comply with some external policies, where the threat are supposed to be
> dealt
> 
> Dino Esposito
> 
> Pope, Nick wrote:
>> 
>> Stefan,
>> 
>> 
>> 
>> I prefer the alternative as this makes it clearer to me what is the
>> threat.
>> 
>> 
>> 
>> Nick
>> 
>> 
>> 
>> -----Original Message-----
>> *From:* Stefan Santesson [mailto:[e-mail]]
>> *Sent:* 29 September 2009 10:28
>> *To:* Pope, Nick; [e-mail]; pkix
>> *Subject:* Re: [pkix] I-D Action:draft-ietf-pkix-rfc3161-update-06.txt
>> 
>> 
>> 
>> I'm sorry to say that I don't think we are doing progress.
>> Rather the text seems to be getting more and more confusing to an
>> outside reader.
>> 
>> I think the following would be much more helpful to a reader.
>> 
>> ESSCertID provide means based on the SHA-1 hash algorithm for
>> identifying the certificate used to verify the signature on a time
>> stamp. The use of ESSCerIDv2 aims to enable implementers to comply
>> with policies that require phasing out all uses of the SHA-1
>> algorithm. Upgrading the hash function through ESSCertIDv2 address
>> some theoretical scenarios involving a trusted malicious CA. Threats
>> caused by a trusted malicious CA is however out of scope of this
>> specification.
>> 
>> Alternatively this (more elaborate):
>> 
>> ESSCertID provide means based on the SHA-1 hash algorithm for
>> identifying the certificate used to verify the signature on a time
>> stamp. The use of ESSCerIDv2 aims to enable implementers to comply
>> with policies that require phasing out all uses of the SHA-1
>> algorithm. Upgrading the hash function through ESSCertIDv2 address a
>> theoretical scenario involving a trusted malicious CA issuing a false
>> but trusted TSA certificate where the hash of the false certificate
>> is identical to the hash of the original TSA certificate. Threats
>> caused by a trusted malicious CA is however out of scope of this
>> specification.
>> 
>> 
>> 
>> Note that the trusted malicious CA have another signing key (unless it
>> has compromised the good CA), thus will produce a totally different
>> signature on the false TSA certificate.
>> We are thus talking about a CA succeeding in creating a SHA-1
>> collision despite a totally different signature value in the end of
>> the certificate. And all this effort just to convince me that the TSA
>> certificate was issued under the wrong policy.....
>> 
>> Seems like a really serious issue...
>> 
>> /Stefan
>> 
>> 
>> On 09-09-29 10:20 AM, "Pope, Nick"<[e-mail]> wrote:
>> 
>> Stefan, Denis,
>> 
>> I would prefer the following, as
>> a) I don't think bringing certificate policy helps clarify things here
>> b) The ESSCertV2 covers the collision threat.
>> 
>> 
>> The introduction of ESSCertID addressed the theoretical threat that
>> a relying party trusts a CA which might possibly malicious
>> generate another TSA certificate. The ESSCertIDV2 addressed the
>> additional theoretical threat that hash of the other TSA
>> certificate is identical to the hash of the original TSA
>> certificate.
>> 
>> The use of ESSCerIDv2 aims to enable implementers to comply with
>> policies that require phasing out all uses of the SHA-1 algorithm.
>> In case SHA-1 might in the future exhibit hash collisions, the
>> introduction of ESSCertIDv2 allows the use of stronger hash
>> functions making the generation of such other certificates close
>> to impossible.
>> 
>> 
>> 
>> 
>> -----Original Message-----
>> *From:* Stefan Santesson [mailto:[e-mail]]
>> *Sent:* 29 September 2009 09:06
>> *To:* [e-mail]; pkix; Pope, Nick
>> *Subject:* Re: [pkix] I-D Action:draft-ietf-pkix-rfc3161-update-06.txt
>> 
>> Denis,
>> 
>> I have no problem with the disposition change.
>> 
>> The problem is that you insist on painting this in words making it
>> look like as a serious security issue.
>> The rest of us seems to agree that it isn't a realistic threat.
>> 
>> To me, the picture you paint sounds very much like reinforcing the
>> door of a vault that has lost it's roof.
>> It does help if the thief tries to enter through the door, but does
>> not prevent anyone from climbing over the wall.
>> 
>> Likewise, for the unlikely event that you will trust a malicious CA,
>> ESSCertIDv2 will probably not save you either.
>> That CA can simply fool you in so many other ways where ESSCertIDv2
>> will not help.
>> 
>> 
>> I think we have to choices. Either we define trusted but malicious CA
>> as a threat that is either:
>> 
>> 1) in scope; or
>> 2) out of scope,
>> 
>> of threats that need to be addressed.
>> 
>> If the trusted malicious CA is in scope, then we need to address all
>> things that a trusted malicious CA can do and come up with
>> recommendations for each one.
>> If the trusted malicious CA is out of scope, then we should say so and
>> not pretend that we fixed the bad CA scenario by mending one of the
>> holes in the broken barrel.
>> 
>> I hope we can all agree that it would be infeasible to define the
>> trusted malicious CA scenario as an "in scope" threat.
>> 
>> /Stefan
>> 
>> 
>> On 09-09-25 4:57 PM, "Denis Pinkas"<[e-mail]> wrote:
>> Stefan and Nick,
>> 
>> I have problems with the second and third paragraphs from section 6.
>> 
>> The text starts with :
>> 
>> The introduction of ESSCertIDv2 addresses the theoretical threat .
>> 
>> The threat already existed with ESSCertID which means that the
>> sentence is incorrect.
>> 
>> I propose the following:
>> 
>> The introduction of ESSCertID addressed the theoretical threat that
>> a relying party trusts a CA which might possibly malicious
>> generate another TSA certificate where the hash of the other TSA
>> certificate is identical to the hash of the original TSA
>> certificate. Since the TSA certificate indicates the
>> Certification Policy under which the TSA key pair has been
>> handled, it is important that the TSA can reliably indicate under
>> which Certification Policy the certificate has been obtained.
>> 
>> The use of ESSCerIDv2 aims to enable implementers to comply with
>> policies that require phasing out all uses of the SHA-1 algorithm.
>> In case SHA-1 might in the future exhibit hash collisions, the
>> introduction of ESSCertIDv2 allows the use of stronger hash
>> functions making the generation of such other certificates close
>> to impossible.
>> 
>> 
>> Denis
>> 
>> ----------
>> 
>> *De :* pkix-bounces <mailto :[e-mail]>
>> 
>> *À :* i-d-announce <mailto :[e-mail]>
>> 
>> *Date :* 2009-09-19, 10:45:01
>> 
>> *Sujet :* [pkix] I-D Action:draft-ietf-pkix-rfc3161-update-06.txt
>> 
>> 
>> 
>> 
>> 
>> *_Pièce(s) Jointe(s) au message original :
>> _*
>> (1). draft-ietf-pkix-rfc3161-update-06.txt
>> 
>> 
>> 
>> Title : ESSCertIDv2 update for RFC 3161
>> Author(s) : S. Santesson, N. Pope
>> Filename : draft-ietf-pkix-rfc3161-update-06.txt
>> Pages : 8
>> Date : 2009-09-19
>> 
>> This document updates RFC 3161 [RFC3161]. It allows the use of
>> ESSCertIDv2 defined in RFC 5035 [ESSV2] to specify the hash of a
>> signer certificate when the hash is calculated with a function other
>> than SHA-1 [SHA1].
>> 
>> A URL for this Internet-Draft is:
>> http://www.ietf.org/internet-drafts/draft-ietf-pkix-rfc3161-update-06.txt
<mailto:[e-mail]> 
>> 
>> Internet-Drafts are also available by anonymous FTP at:
>> ftp://ftp.ietf.org/internet-drafts/
>> 
>> Below is the data which will enable a MIME compliant mail reader
>> implementation to automatically retrieve the ASCII version of the
>> Internet-Draft.
>> 
>> 
>> _____________________ next part ______________________
>> 
>> _______________________________________________
>> pkix mailing list
>> [e-mail] <mailto: [e-mail]>
>> https://www.ietf.org/mailman/listinfo/pkix
>> 
>> 
>> ------------------------------------------------------------------------
>> 
>> 
>> _______________________________________________
>> pkix mailing list
>> [e-mail]
>> https://www.ietf.org/mailman/listinfo/pkix
>> 
>>------------------------------------------------------------------------
>> 
>> _______________________________________________
>> pkix mailing list
>> [e-mail]
>> https://www.ietf.org/mailman/listinfo/pkix
>> 
>> *_/Consider the environment before printing this mail./_*
>> 
>> */"Thales e-Security Limited is incorporated in England and Waleswith>
>> company registration number 2518805. Its registered office is located
>> at 2 Dashwood Lang Road, The Bourne Business Park, Addlestone, Nr.
>> Weybridge, Surrey KT15 2NX./*
>> 
>> */The information contained in this e-mail is confidential. It may
>> also be privileged. It is only intended for the stated addressee(s)
>> and access to it by any other person is unauthorised. If you are not
>> an addressee or the intended addressee, you must not disclose, copy,
>> circulate or in any other way use or rely on the information contained
>> in this e-mail. Such unauthorised use may be unlawful. If you have
>> received this e-mail in error please delete it (and all copies) from
>> your system, please also inform us immediately on [phone]
>> or email [e-mail]. Commercial matters detailed
>> or referred to in this e-mail are subject to a written contract signed
>> for and on behalf of Thales e-Security Limited"./*
>> 
>>------------------------------------------------------------------------
>> 
>> _______________________________________________
>> pkix mailing list
>> [e-mail]
>> https://www.ietf.org/mailman/listinfo/pkix
>> 
> _______________________________________________
> pkix mailing list
> [e-mail]
> https://www.ietf.org/mailman/listinfo/pkix


_______________________________________________
pkix mailing list
[e-mail]
https://www.ietf.org/mailman/listinfo/pkix
Consider the environment before printing this mail.

"Thales e-Security Limited is incorporated in England and Wales with company
registration number 2518805. Its registered office is located at 2 Dashwood
Lang Road, The Bourne Business Park, Addlestone, Nr. Weybridge, Surrey KT15
2NX.

The information contained in this e-mail is confidential. It may also be
privileged. It is only intended for the stated addressee(s) and access to it
by any other person is unauthorised. If you are not an addressee or the
intended addressee, you must not disclose, copy, circulate or in any other
way use or rely on the information contained in this e-mail. Such
unauthorised use may be unlawful. If you have received this e-mail in error
please delete it (and all copies) from your system, please also inform us
immediately on [phone] or email [e-mail].
Commercial matters detailed or referred to in this e-mail are subject to a
written contract signed for and on behalf of Thales e-Security Limited". 
_______________________________________________pkix mailing
[e-mail]://www.ietf.org/mailman/listinfo/pkix
<[e-mail]:/www.ietf.org/mailman/listinfo/pkix> 
Consider the environment before printing this mail.

"Thales e-Security Limited is incorporated in England and Wales with company
registration number 2518805. Its registered office is located at 2 Dashwood
Lang Road, The Bourne Business Park, Addlestone, Nr. Weybridge, Surrey KT15
2NX.

The information contained in this e-mail is confidential. It may also be
privileged. It is only intended for the stated addressee(s) and access to it
by any other person is unauthorised. If you are not an addressee or the
intended addressee, you must not disclose, copy, circulate or in any other
way use or rely on the information contained in this e-mail. Such
unauthorised use may be unlawful. If you have received this e-mail in error
please delete it (and all copies) from your system, please also inform us
immediately on [phone] or email [e-mail].
Commercial matters detailed or referred to in this e-mail are subject to a
written contract signed for and on behalf of Thales e-Security Limited".