We’re steadily learning more about the latest Trusted Computing proposals. People have started to grok that building signed boot into UEFI will extend Microsoft’s power over the markets for AV software and other security tools that install around boot time; while ‘Metro’ style apps (i.e. web/tablet/html5 style stuff) could be limited to distribution via the MS app store. Even if users can opt out, most of them won’t. That’s a lot of firms suddenly finding Steve Ballmer’s boot on their jugular.
We’ve also been starting to think about the issues of law enforcement access that arose during the crypto wars and that came to light again with CAs. These issues are even more wicked with trusted boot. If the Turkish government compelled Microsoft to include the Tubitak key in Windows so their intelligence services could do man-in-the-middle attacks on Kurdish MPs’ gmail, then I expect they’ll also tell Microsoft to issue them a UEFI key to authenticate their keylogger malware. Hey, I removed the Tubitak key from my browser, but how do I identify and block all foreign governments’ UEFI keys?
Our Greek colleagues are already a bit cheesed off with Wall Street. How happy will they be if in future they won’t be able to install the security software of their choice on their PCs, but the Turkish secret police will?
I’m curious: does this add security in any way then?
would it be better to have an opt-in instead of opt-out model for secure boot?
UEFI Rule Number 1: Never buy a UEFI system without JTAG support (i.e. an actual JTAG connector on the mother board). Anything that the Manufacturer can flash into memory can be undone by using the same equipment. Read, Modify, re-Write the flash image with you own desired keys.
I foresee an after market device to simplify the whole process: Simply write your desired keys on to a USB dongle, carry it over to the machine being held hostage, plug the JTAG cable into the port on the mother board, insert the other end into the USB port for power, and power up the machine. Problem solved. Go ahead and boot your favourite software.
Ongoing efforts on the Linux side to avoid Microsoft’s control over PC vendors resulting in PCs which can only boot Windows are summarised by Matthew Garrett:
http://mjg59.dreamwidth.org/5552.html
http://mjg59.dreamwidth.org/5850.html
http://mjg59.dreamwidth.org/6503.html
Whilst this proposal isn’t without its drawbacks, it might be enough to take some of the control over what you can and can’t boot away from Microsoft and return it to the user.
Linux will never make it to the home market, fact, so what’s the worry? Any server hardware product should not come with the the secure boot option as they are more likely to change/reinstall/repurpose machines than the average home user. Yes, it may seem unfair Microsoft have an monopoly, does it mean you should selfishly want average home users PC’s to be less secure? No. Microsoft and the hardware vendors should not give any government the keys, and if they do I suspect something similar to the Arab Spring should arise if found to be true.
OS/X is a suitable alternative, oh that hardware it runs on. And you have to use Apple’s Boot Camp to boot Windows without using VM’s. I suspect this will too move to the UEFI with secure boot soon, will there be any fuss over that? very little. Why? Because they’re Apple. They ‘make’ the hardware.
Microsoft have already managed to poison the debate here by using their favoured approach of redefining the meaning of words. What most people would expect from the phrase Trusted Computing would be the user’s preferred quality checked and validated OS with the user’s chosen AV/Anti-SPAM etc loaded on top ie, the user has set it up how they want and no-one can slide in and play with it.
The Microsoft meaning is that only someone with a pre-approved key can create/change the base structure unless mobo manufacturers risk their Win8 compliance rating by allowing users to add/change/remove keys. Oh, and who supplies those special keys to the PC manufacturers?
Ross thanks for bringing the necessary focus on this. We all know what MS are capable of if they are not constantly watched. And for the dimwit going on about OS market shares, how are we making computers less secure by allowing users to bypass this/ add keys?
There were some conversations around this at the Kernel Summit. I believe the Linux Foundation will be issuing its own recommendations on how firmware and operating systems should implement key management for Secure Boot in order to support end-user choice.
AV software…what?
my reading of the uefi spec indicates that only the bootloader is verified & once it has been control is passed to it to load whatever it likes – which would include AV modules.
AV companies don’t write bootloaders & so wouldn’t be affected.
rootkits replace the boot loader and would not be verified meaning the system won’t boot if infected. (if you have secure boot enabled) which is where the security feature comes in.
Jay,
I’ve been using Linux at home, pretty much exclusively, for over 10 years. So you are sadly misinformed. I do Linux systems work for a living, so it was more convenient. I stayed for the productivity gains. All the tools I need come with the OS, I don’t have to stay fully patched (though I usually am, because it’s dead easy) to avoid the latest drive-by download exploit, etc.
My productivity is *at least* 2X what it would be on Windows, and the hardware demands are lower, for a faster machine. I get more system life time for my outlay.
There are too many other advantages to go into here, all of which I’m supposed to abandon. To support a return to monopolistic business practices?
Here’s a tip: what you (and/or your employer) do with computers does not constitute the whole of computing. Linux covers a large space; embedded, business (and home!) laptops/desktops, generic servers, storage arrays, and Top500 supercomputers.
Nor is your security argument valid. This will not protect users from way over 99% of exploits currently seen in the wild. Some malware (the TDL rootkit family comes to mind) gets security industry press because of the sheer level of cunning involved, and would be defended against. But it simply isn’t necessary to go to those lengths to completely compromise the *vast* majority of home-user (or corporate, for that matter) Windows systems.
The numbers don’t add up; this *may* be about marketing, but it most certainly is *not* about security. Given the history of Microsoft, I lean toward vendor lock-in. That’s what system software vendors *do*, for the most part. It makes the spreadsheets come out nicer.
What profound nonsense. Microsoft doesn’t issue UEFI keys. Anyone can sign a UEFI file. It just needs to be signed.
So, is it better to be able to look at who singed your code and know that the Turkish governmet signed something that is running on your computer or to have nothing be signed and not know where the hell it came from.
I guess is it more fun (and profitable) to take uninformed cheap shots than it is to understand the issues and give people useful information.
Most Smartphones already have secure boot enabled. The keys belong to the OEM (eg. Motorola or Sony) and NOT the OS manufacturer (eg. Google or Microsoft) unless both belong to the same company (eg. Apple and RIM).
This does not prevent people from jailbreaking those devices, and recently OEMs that use secure boot started allowing users to unlock the bootloaders. Users do lose some functionality such as DRM keys when they officially unlock their devices, which is to be expected.
There is nothing that suggests that things will be different on PCs. The OEMs and not Microsoft or RedHat or Ubuntu will maintain the root keys. Users will just have to make sure their bootloaders can be unlocked before buying their computers.
Is there any particular reason to pick on TUBITAK? It really *is* an independent research institute although, unfortunately, the present government is slowly and steadily undermining its independence.
TUBITAK does have a national cryptographic research division (UEKAE), and does work for military, but it does not mean that it is controlled by the military or the intelligence.
I am also very curious to know what is special about Turkey, compared with say Japan, US, Korea, Mexico, Serbia or Tunisia and other countries that have government-affiliated root certs in Windows, and in all major Browsers?
It probably makes a good story, but unless there is any evidence that suggests that Turkish government/military/intelligence is using these certs for eavesdropping on communications, then it is just fluff that obscures the fundamental problems of the CA system.
Iran did not need its government CA to eavesdrop on its citizens, for instance. It just used hacked DigiNotar certificates instead (which ironically also hosted the root keys of the government of The Netherlands). The misuse still got detected and word got out even though the Iranian Internet is very strictly controlled.
I would love to know about any evidence of the misuse of TUBITAK certs by the way. I am not ruling it out by any means. There is very little awareness/respect for user privacy and a growing tendency to monitor/censor/control communications in Turkey these past years. It would be very good to know if they are using official TUBITAK (or any other Turkish CA) certs for this purpose.
Here’s the latest from the Linux Foundation. They, and others, discuss how secure boot might be implemented in ways that might not damage the free and open source software ecology.
[citation needed]