On Fri, 2008-04-18 at 13:22 +0200, Robert Millan wrote: > Hi Laurent, > > The problem with these TPM chips is that they have the hidden purpose of > restricting you as user. Despite that you paid for the hardware and are its > owner, the chip will never give you its master key.
Sorry, but this message is confusing me. Having the TPM in my machine act as a cryptographic proxy on my behalf is the entire point of the TPM: if the software stack has access to the SRK then attackers would prefer to attack dead swap space or temp files rather than the TPM itself. > The idea behind this is that you can be coerced into accepting that someone > else can spy on your computer (they call it "remote attestation"). When > enough users accept this form of blackmail, it will become impossible to > resist to it in practice. And this is the really confusing part. How can someone else spy on my computer because of my TPM? I can *voluntarily* enter into a remote attestation system, but to do that I would need to tell my peers the public key I will be using to sign the attestations; if I was so inclined, I could choose any key that I like for this purpose, and instruct the software on my machine to get the unencrypted PCRs from my TPM, modify their values as I saw fit, and sign that configuration instead. Even if the software that runs the remote attestation is honest (say, because I'm running some Windows-based scheme that I can't easily change), I can still elect to boot into Linux, authenticate to the TPM with the owner password, and ask it to perform whatever operations I want with whatever PCR configuration I want. > For these reasons, I'd like to encourage you to consider the ethical > implications of using and supporting this technology, and look for > alternatives that would satisfy whatever needs you had in it (I'd welcome > some discussion about that, to see how GRUB can help). -- -Julian Blake Kongslie <[EMAIL PROTECTED]> If this is a mailing list, please CC me on replies. vim: set ft=text : _______________________________________________ Grub-devel mailing list Grub-devel@gnu.org http://lists.gnu.org/mailman/listinfo/grub-devel
