So, just to be clear: Imagine a scenario with two devices - one I made (I know the keys and code on BL1) and another one that some malicious user cloned (he signed with his own keys). My device will have a Root of Trust in BL1 based on my hardware and the keys I own. The second device also has a BL1 but that image was signed by someone I don't trust. In the end, both devices will boot up successfully because they are based on each individual Chain of Trust but there's no way a third party (i.e. remote attestation server) can know the difference between the malicious device and my device solely relying on Verified Boot, right?
Tue, Sep 10
Mon, Sep 9
Verified boot in itself already proves the boot integrity of all firmware images from BL1 up to BL33.
BL1 is the root of trust and cannot be tampered with, as it is hardware-protected. All other BL image are signed and their signature is verified before they get executed: BL1 verifies the signature of BL2, and BL2 does the same for all subsequent BL3x images. If one of the signatures is invalid then TF-A refuses to execute the corresponding image. This is treated as a fatal error that the firmware cannot recover from and the platform will typically panic in this case.
Sorry, I completely missed your point at first!
Fri, Sep 6
Thu, Sep 5
Are you thinking something similar to measured boot ?
The TF-A implements what is called verified boot. Found a good description for difference between verified and measured boot here:
Aug 7 2019
Jun 24 2019
Reopening so it can be answered
It’s better to create a task from scratch than edit the example task. Added back in TF-A project so it is shown on the project workbook with your question.
Jun 16 2019
May 31 2019
May 28 2019
May 21 2019
Thanks guys! The dmbish() is not a huge deal. Just get a little nervous when i see barriers and don't completely understand why it is there. :)
Ah, You are right. Having taken a look at it again, yes, the SP-> SPM communication is register based and this spm_response_add() is invoked by SPM to push to a buffer within EL3 (its not a shared buffer between different ELs). I suspect the shared buffer primitives were written with shared buffer scenario in mind and the current prototype implementation does not optimize it for the case when the buffer is within EL3.
Thanks Paul, Soby.
spm_response_*() currently cannot invoked by any secure partition since the responses array is in EL3 space. Is this not the case ? or is it the expectation that the responses array will be mapped to secure EL0 some time in the future? I don't see how a secure partition can invoke spm_response_* other than through an SMC, in which case we are already in EL3 context and dont require the dmbish(), as Paul pointed. I understand your argument for sprt_queue_*, since they are invoked by EL3 and the secure partition.
Who is the lockless reader for spm_response_add() and spm_response_get()?
Apologies for the delay. I've had a look into the spm_buffers.c file and I understand your query is around the dmbish() that's performed shortly before releasing the lock in both spm_response_add() and spm_response_get().
May 13 2019
May 8 2019
Thanks. Missed the lockless reader of the queue. Who is the lockless reader for spm_response_add() and spm_response_get()?
A secure partition may be reading from the queue from CPU1 at the same time as CPU0 pushes a new request. If this happens, it is needed that the stores are seen in the same order by all observers of the system.