On Mon, 2018-03-12 at 12:58 +0200, Jarkko Sakkinen wrote: > On Sat, 2018-03-10 at 14:13 -0800, James Bottomley wrote: > > > > By now, everybody knows we have a problem with the TPM2_RS_PW easy > > button on TPM2 in that transactions on the TPM bus can be > > intercepted > > and altered. The way to fix this is to use real sessions for HMAC > > capabilities to ensure integrity and to use parameter and response > > encryption to ensure confidentiality of the data flowing over the > > TPM > > bus. > > > > This patch series is about adding a simple API which can ensure the > > above properties as a layered addition to the existing TPM handling > > code. This series now includes protections for PCR extend, getting > > random numbers from the TPM and data sealing and unsealing. It > > therefore eliminates all uses of TPM2_RS_PW in the kernel and adds > > encryption protection to sensitive data flowing into and out of the > > TPM. > > > > This series is also dependent on additions to the crypto subsystem > > to > > fix problems in the elliptic curve key handling and add the Cipher > > FeedBack encryption scheme: > > > > https://marc.info/?l=linux-crypto-vger&m=151994371015475 > > > > In the third version I've added data sealing and unsealing > > protection, > > apart from one API based problem which means that the way trusted > > keys > > were protected it's not currently possible to HMAC protect an > > authority > > that comes with a policy, so the API will have to be extended to > > fix > > that case > > > > I've verified this using the test suite in the last patch on a VM > > connected to a tpm2 emulator. I also instrumented the emulator to > > make > > sure the sensitive data was properly encrypted. > > > > James > > 1. Can I ignore v2 and just review/test this version? I haven't even > peeked into v2 yet. Yes, v3 is a more complete version of v2 with a couple of sessions API additions. I think the way I'm going to fix the trusted key policy problem is to move it back into the kernel for the simple PCR lock policy (which will make changing from 1.2 to 2.0 seamless because the external Key API will then become the same) so the kernel gets the missing TPM nonce and can then do TPM2_PolicyAuthValue. User generated policy sessions for trusted keys are very flexible but also a hugely bad idea for consumers because it's so different from the way 1.2 works and it means now the user has to exercise a TPM API to produce the policy sessions. Longer term, I think having a particular trusted key represent a policy session which can then be attached to a different trusted key representing the blob is the best idea because we can expose the policy build up via the trusted key API and keep all the TPM nastiness inside the kernel. > 2. Do you know in which kernel version will the crypto additions > land? They're here: https://git.kernel.org/pub/scm/linux/kernel/git/herbert/cryptodev-2.6.git/log/ So I'd guess next merge window. You can do what we do in SCSI and create a "postmerge" branch based on the cryptodev one (we often have SCSI stuff with block tree precursors). The way I run it is that I don't send the merge window pull request until I see the merge-base against Linus master move to the base of the patches (meaning all the precursors are upstream). > /Jarkko >
