Post AXb8LQIr6kgdbfT0LY by c0dec0dec0de@hachyderm.io
(DIR) More posts by c0dec0dec0de@hachyderm.io
(DIR) Post #AXZvNxSOGn4E8t6Pei by mjg59@nondeterministic.computer
2023-07-11T07:58:39Z
1 likes, 2 repeats
Ok why is it hard to implement a hardware root of trust? https://mjg59.dreamwidth.org/66907.html
(DIR) Post #AXaLYh3TA9I5coVfqS by hipolito@social.restless.systems
2023-07-11T12:51:26Z
0 likes, 0 repeats
@mjg59Thanks for writing it, it was quite clear. I saw a comment in the Orange Website for the Purism root of trust method.
(DIR) Post #AXaSyenCtbB9oiEMmu by th@social.v.st
2023-07-11T14:14:39Z
0 likes, 0 repeats
@mjg59 I *finally* encountered a platform that actually used ACM to initialize the TPM from locality 3, reducing the risk of "how do we know the CPU has Boot Guard enabled?"
(DIR) Post #AXaVXsviWdkvhs8Yd6 by ejrowley@mastodon.green
2023-07-11T14:43:14Z
0 likes, 0 repeats
@mjg59 I agree this is a great problem for the industry to address. What you might find interesting is the "binary transparency" example Google provided for their Trillian distributed ledger. It doesn't solve this problem at the CPU level, but it is an interesting approach for firmware, and would help with the compromised signing key example you cited (Realtek/Stuxnet). It's the same technology behind much of Certificate Transparency. See here: https://github.com/google/trillian-examples/tree/master/binary_transparency/firmware
(DIR) Post #AXb0mJEGiYDfqud9ge by ljrk@todon.eu
2023-07-11T20:33:22Z
0 likes, 0 repeats
@mjg59 I love your posts, not only because they're precisely what I'm interested in but also because they're written extremely approachable and als always have a few chuckles ready. Thanks for taking your time :3
(DIR) Post #AXb8LQIr6kgdbfT0LY by c0dec0dec0de@hachyderm.io
2023-07-11T21:58:21Z
0 likes, 0 repeats
@mjg59 RET_TRUST_ME_BRO