On Sat, 22 Jun 2002, Lucky Green wrote:

I must confess that after reading the paper I am quite relieved to finally have solid confirmation that at least one other person has realized (outside the authors and proponents of the bill) that the Hollings bill, while failing to mention TCPA anywhere in the text of the bill, was written with the specific technology provided by the TCPA in mind for the purpose of mandating the inclusion of this technology in all future general-purpose computing platforms, now that the technology has been tested, is ready to ship, and the BIOS vendors are on side.

A touch hand wavy, but interesting. (and thank you to JY for the pointer.)

"trusted" here means that the members of the TCPA trust that the TPM will make it near impossible for the owner of that motherboard to access supervisor mode on the CPU without their knowledge, they trust that the TPM will enable them to determine remotely if the customer has a kernel-level debugger loaded, and they trust that the TPM will prevent a user from bypassing OS protections by installing custom PCI cards to read out memory directly via DMA without going through the CPU.

I don't see how they expect this to work. We've already got cheap rip off motherboards, who's gonna stop cheap rip off TPM's that ain't really T? I think it moves the game into a smaller field where the players all have some bucks to begin with, but somebody will create a "TPM" that looks like the real thing, but runs cypherpunk code just fine.

1) the CEO's of said computer companies are utterly unaware of a major strategic initiative their staff has been diligently executing for about 3 years, in the case of the principals in the TCPA, such as Intel, Compaq, HP, and Microsoft, several years longer.

2) the CEO's wrote this open letter as part of a deliberate "good cop, bad cop" ploy, feigning opposition to DRM in general computing platforms to pull the wool over the public's eye for hopefully long enough to achieve widespread deployment of the mother of all DRM solution in the market place.

3) some people think DRM will work and some people don't, and they all work at the same company. Anyone who can comprehend the physical reality of computers can see DRM can't possibly work. Unfortunatly, that's a minorty of the human population. I think the CEO's may actually have a clue, but if there's money to be made from suckers, why not!!?? Well, I know why not, and so do you all. But I don't think mandated "Fritz" chips will fly - and it's simple economics. Logic will never work :-) Patience, persistence, truth, Dr. mike

Mike wrote quoting Lucky:

...

"trusted" here means that the members of the TCPA trust that the TPM will make it near impossible for the owner of that motherboard to access supervisor mode on the CPU without their knowledge, they trust that the TPM will enable them to determine remotely if the customer has a kernel-level debugger loaded, and they trust that the TPM will prevent a user from bypassing OS protections by installing custom PCI cards to read out memory directly via DMA without going through the CPU.

I don't see how they expect this to work. We've already got cheap rip off motherboards, who's gonna stop cheap rip off TPM's that ain't really T? I think it moves the game into a smaller field where the players all have some bucks to begin with, but somebody will create a "TPM" that looks like the real thing, but runs cypherpunk code just fine.

I agree with your assertion that TPM's can't prevent DRM from being broken. Nor is this the intent of introducing TPM's. The vendors have realized that they have to raise the technical bar only so high to keep those most inclined to break their systems (i.e. 16-year old Norwegians) from doing so. Those that have the knowledge and resources to break TCPA systems either won't have the time because they are engaged in gainful employment, won't be willing to take the risk, because they have accumulated sufficient material possessions to be unwilling to risk losing their possessions, not to mention their freedom, in litigation, or will break the security for their own gain, but won't release the crack to the public. Criminal enterprise falls into the latter category. The content vendors, which in this case includes the operating system and application vendors, dislike, but can live with, major criminal enterprise being the only other party to have unfettered access, since criminal enterprise is just another competitor in the market place. Most business models can survive another competitor. Where business models threaten to collapse is when the marginal cost of an illegal copy goes to zero and the public at large can obtain your goods without payment. I don't know if the TCPA's efforts will prevent this, but in the process of trying to achieve this objective, the average computers users, and even many advanced computer users, will find themselves in a new relationship with their PC: that of a pure consumer, with only the choices available to them the what the 180 TCPA's members digital signatures permit. Cloning TPM's is difficult, though not impossible. Note that all TPM's unique initial internal device keys are signed at time of manufacture by a derivative of the TCPA master key. Unless you are one of the well-known chipset or BIOS manufacturers, you can't get your TPM products signed. It is theoretically possible, though far from easy, to clone an entire TPM, keys and all. However, the moment those fake TPM's show up in the market place, their keys will simply be listed in the next CRL update. And if your OS and TPM's miss a few CRL updates, your commercial OS and all your applications will stop working. As might in the future your video card, your PCI cards, your hard drive, and your peripherals. You can try to hack around the code in the OS or firmware that performs the checks, as long as you are willing to operate your machine permanently off the Net from then on, because your system will fail the remote integrity checks, but given that this and other security relevant code inside the OS and applications are 3DES encrypted and are only decrypted inside the TPM, you can't just read the object code from disk, but get to first microprobe the decrypted op codes off the bus before taking a debugger to the code. Not a trivial task at today's PC bus speeds. Nor can you get too aggressive with the hacks, since your Fritz may simply flush the keys and leave you with a bunch of 3DES encrypted op codes and no corresponding decryption keys. Reverse engineering turns pretty dim at that point. None of these obstacles are impossible to overcome, but not by Joe Computer User, not by even the most talented 16-year old hacker, and not even by many folks in the field. Sure, I know some that could overcome it, but they may not be willing to do the time for what by then will be a crime. Come to think of it, doing so already is a crime. --Lucky Green --------------------------------------------------------------------- The Cryptography Mailing List Unsubscribe by sending "unsubscribe cryptography" to majordomo@wasabisystems.com