Re: [Trustedjava-support] Encrypting the private part of an RSA key

[Jonathan M. <jon...@cm...>]

You might also look at IBM's software implementation to see how they
implemented the internals of the TPM in this regard:

http://ibmswtpm.sourceforge.net/

-Jon


On Fri, Jan 13, 2012 at 10:14 AM,  <Fed...@ff...> wrote:
> Hi again,
>
> Maybe I found the answer myself, so I will report it just in case anyone
> else ever wondered or someone who already knows can correct me if I am
> wrong.
>
> After going through the source code of the jTSS I realized that the wrapping
> of a key is done inside the TPM, os I went to look at the TPM specs, in
> particular the data structure part, and there it says the an asymmetric key
> is stored so that the private part cannot be bigger than 151 bytes
> (TPM_STORE_PRIVKEY).
>
> This is possible because it is not the private exponent that is stored as
> private key (which is usually nearly as big as the modulus itself), but one
> of the two prime factor of the modulus, and 151 bytes CAN be encrypted by a
> 2048 RSA public key in one block.
>
>
>
> Federico
>
>
>
> Fra: Fed...@ff... [mailto:Fed...@ff...]
> Sendt: 12. januar 2012 09:18
> Til: sha...@ya...
> Kopi: tru...@li...
>
>
> Emne: Re: [Trustedjava-support] Encrypting the private part of an RSA key
>
>
>
> Hi,
>
> Thanks for your pointer, but I had already read the book, and that is what
> confused me.
>
> In chapter 2 “Secure Storage – Storing asymmetric keys” it says that
> asymmetric keys are encrypted under a 2048 RSA key and the format used is
> the PKSC#1 v2.0. My question was about how this was done in practice since a
> public 2048 RSA key can encrypt less than 256 bytes and a RSA private key is
> usually at least as big as the modulus, i.e. 256 bytes. I read also the
>  PKSC#1 v2.0 (not line by line, but I had a close look), and I could not
> find any specific guidelines for encrypting private keys, only to format and
> encode them.
>
> In other words, if at some point the TPM produces an AIK, and this must be
> stored outside the TPM protected by the SRK key, how is this done in
> practice?
>
> Or is the AIK key encrypted with a symmetric key in turn encrypted with the
> SRK public key?
>
> Also a pointer to the source code where this is implemented would be fine,
> if possible.
>
>
>
> Thanks again, and sorry if the question is trivial, but I am clearly missing
> some probably stupid detail here, and I cannot seem to figure out what it
> is.
>
>
>
> Federico
>
>
>
> Fra: Shakir Ullah shah [mailto:sha...@ya...]
> Sendt: 12. januar 2012 05:59
> Til: Mancini, Federico
> Emne: Re: [Trustedjava-support] Encrypting the private part of an RSA key
>
>
>
> Hi
>
>
>
> You can find answer to your question by reading chapter 2 of
> IBM.Press.A.Practical.Guide.to.Trusted.Computing.Jan.2008 book.
>
>
>
> Regards
>
> Shakir
>
>
>
> From: "Fed...@ff..." <Fed...@ff...>
> To: tru...@li...
> Sent: Wednesday, January 11, 2012 8:15 PM
> Subject: [Trustedjava-support] Encrypting the private part of an RSA key
>
>
>
> Hi,
>
> I am new here and I have only recently started reading up on trusted
> computing and TPM for a project I am part of.
>
> I think I got all the basics, but now that I am entering more technical
> details, I find all my books and sources are not enough any more.
>
> That is why I have starting looking at your implementation of the TSS (which
> I plan to use soon to implement some ideas) and  why I am posting here,
> since whoever implemented something like this, must have a very good and
> practical knowledge of the specifications J
>
> The question I have is about how asymmetric keys are stored securely outside
> the TPM.
>
> As far as I understand, AIK or Storage keys are asymmetric pairs of keys and
> are stored outside the TPM in a secure way protected by their parent key.
>
> I guess this means that the private part of these keys must be encrypted,
> and in particular it should be encrypted with the SRK or
>
> other parent storage keys, which are also asymmetric. So how is this done?
> By encrypting the private part of a key with the public part of its parent
> key (possibly the SRK itself)?
>
> If so, isn’t a private key too big to fit in one RSA encrypted block? Does
> it have to be split in smaller blocks first? If so how (is there a standard)
> ?
>
> Reading your Javadoc I found at least two methods that return an encrypted
> private key (TspCreateWrapKey_Internal), so I hope you can help me out with
> this.
>
>
>
> This question is bugging me since in the book “A practical guide to trusted
> computing” there is a whole section dedicated to how big a symmetric key can
> be to be encrypted by a binding (asymmetric) key,
>
> but there does not seem to be a problem to encrypt potentially big private
> keys instead.
>
>
>
> Thanks for any answer!
>
>
>
> Federico Mancini
>
>
>
>
> ------------------------------------------------------------------------------
> Ridiculously easy VDI. With Citrix VDI-in-a-Box, you don't need a complex
> infrastructure or vast IT resources to deliver seamless, secure access to
> virtual desktops. With this all-in-one solution, easily deploy virtual
> desktops for less than the cost of PCs and save 60% on VDI infrastructure
> costs. Try it free! http://p.sf.net/sfu/Citrix-VDIinabox
> _______________________________________________
> Trustedjava-support mailing list
> Tru...@li...
> https://lists.sourceforge.net/lists/listinfo/trustedjava-support
>
>
> ------------------------------------------------------------------------------
> RSA(R) Conference 2012
> Mar 27 - Feb 2
> Save $400 by Jan. 27
> Register now!
> http://p.sf.net/sfu/rsa-sfdev2dev2
> _______________________________________________
> Trustedjava-support mailing list
> Tru...@li...
> https://lists.sourceforge.net/lists/listinfo/trustedjava-support
>