On Thu, Mar 26, 2020 at 2:28 PM Matthew Garrett <mjg59@xxxxxxxxxx> wrote: > > On Thu, Mar 26, 2020 at 2:07 PM Andy Lutomirski <luto@xxxxxxxxxxxxxx> wrote: > > > On Mar 26, 2020, at 1:40 PM, Matthew Garrett <mjg59@xxxxxxxxxx> wrote: > > > > > > On Thu, Mar 26, 2020 at 1:33 PM Andy Lutomirski <luto@xxxxxxxxxxxxxx> wrote: > > >> As a straw-man approach: make the rule that we never call EFI after secure launch. Instead we write out any firmware variables that we want to change on disk somewhere. When we want to commit those changes, we reboot, commit the changes, and re-launch. Or we deactivate the kernel kexec-style, seal the image against PCRs, blow away PCRs, call EFI, relaunch, unseal the PCRs, and continue on our merry way. > > > > > > That breaks the memory overwrite protection code, where a variable is > > > set at boot and cleared on a controlled reboot. > > > > Can you elaborate? I’m not familiar with this. > > https://trustedcomputinggroup.org/wp-content/uploads/TCG_PlatformResetAttackMitigationSpecification_1.10_published.pdf > - you want to protect in-memory secrets from a physically present > attacker hitting the reset button, booting something else and just > dumping RAM. This is avoided by setting a variable at boot time (in > the boot stub), and then clearing it on reboot once the secrets have > been cleared from RAM. If the variable isn't cleared, the firmware > overwrites all RAM contents before booting anything else. I admit my information is rather dated, but I'm pretty sure that at least some and possibly all TXT implementations solve this more directly. In particular, as I understand it, when you TXT-launch anything, a nonvolatile flag in the chipset is set. On reboot, the chipset will not allow access to memory *at all* until an authenticated code module wipes memory and clears that flag. If your computer advertises TXT support but is missing that ACM, you are SOL. I learned about this when I bricked my old Lenovo laptop. As far as I can tell, the flag was set, but the Lenovo BIOS didn't know how to wipe memory. Whoops! > > > > As for the second approach - how would we > > > verify that the EFI code hadn't modified any user pages? Those > > > wouldn't be measured during the second secure launch. If we're calling > > > the code at runtime then I think we need to assert that it's trusted. > > > > Maybe you’re misunderstanding my suggestion. I’m suggesting that we hibernate the whole running system to memory (more like kexec jump than hibernate) and authenticated-encrypt the whole thing (including user memory) with a PCR-sealed key. We jump to a stub that zaps PCRs does EFI calls. Then we re-launch and decrypt memory. > > When you say "re-launch", you mean perform a second secure launch? I > think that would work, as long as we could reconstruct an identical > state to ensure that the PCR17 values matched - and that seems like a > hard problem. Exactly. I would hope that performing a second secure launch would reproduce the same post-launch PCRs as the first launch. If the kernel were wise enough to record all PCR extensions, it could replay them. (I can imagine an alternate universe in which the PCR extension used a more clever algorithm that allowed log-time fast forwarding. As far as I know, this is not currently the case.) In any case, I'm kind of with Daniel here. We survived for quite a long time without EFI variables at all. The ability to write them is nice, and we certainly need some way, however awkward, to write them on rare occasions, but I don't think we really need painless runtime writes to EFI variables.
