Hello Robert,

first I want to thank you again, cause you can't imagine how helpful the few sentences you wrote were. I had lots of puzzle-elements, but couldn't find the way to put them together :)

I also came to the conclusion that encrypting the token-db would be stupid. The whole system is a cryptography system, and I remove modularity from that system if I encrypt the software tokens / the db used by it. I guess the system already includes all the mechanisms I wanted to achieve with my idea. Now I'd like to follow your idea, but could you please tell me if I'm right? The idea is:

Guessed facts: NSS uses a standart Softtoken "pluged" in a logical slot, controled by the module-libraries that are installed with / already imported in Firefox. Access "pin-string" to that token is by default "NULL", or the master password if it is set. Inside that token, I can find the key used to encrypt the Logins by mozStorage. mozStorage uses SecretDecoderRing.c to access that keys and encrypt data with the methods EncryptString(), and Encrypt() before it is stored in the db3-sqliteDB.

Guessed "mission": import opencryptoki-Module that accesses my HW-Slot with the pluged in HW-TPM. Then modify SecretDecoderRing inside the NSS xpcom-module in a way it accesses the other token (the TPM one) instead of the standard Software one.
Ist that right?
How is multi-user-token-using implemented? Does every user have his own Softtoken? I don't know yet what structure I have to recreate inside the TPM, in order to keep the functionalities of the Softtoken(s?). If I have a distinct token for each FF-User, and only one TPM-token for all, how can I manage to keep that multi-user architecture?

For the moment I haven't been able to import the cryptographic module / opencryptoki libraries. There's another thread about that in this newsgroup:
Subject: PKCS#11 Module for TPM availiable
Date: Mon, 6 Jul 2009 07:41:22 -0700 (PDT)


Thanks a lot!

Marc


"Robert Relyea" <[email protected]> schrieb im Newsbeitrag news:[email protected]...
--
dev-tech-crypto mailing list
[email protected]
https://lists.mozilla.org/listinfo/dev-tech-crypto

Reply via email to