Announcement

Collapse
No announcement yet.

AMD Publishes Cryptographic Coprocessor Linux Code

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • AMD Publishes Cryptographic Coprocessor Linux Code

    Phoronix: AMD Publishes Cryptographic Coprocessor Linux Code

    AMD has just published a new set of Linux kernel patches, revealing Linux support for a Cryptographic Coprocessor (AMD CCP)...

    Phoronix, Linux Hardware Reviews, Linux hardware benchmarks, Linux server benchmarks, Linux benchmarking, Desktop Linux, Linux performance, Open Source graphics, Linux How To, Ubuntu benchmarks, Ubuntu hardware, Phoronix Test Suite

  • #2
    But who's going to trust an encryption black box these days? Particularly from a US company?

    Even Intel's AES instructions are dubious; you can verify they encrypt properly in your tests, but you can't verify there isn't some hidden switch in a configuration register or undocumented instruction somewhere which will make them spew out plaintext instead.

    Comment


    • #3
      AMD Publishes Cryptographic Coprocessor Linux Code
      Originally posted by movieman View Post
      But who's going to trust an encryption black box these days? Particularly from a US company?
      Need some reading comprehension there?

      Comment


      • #4
        Originally posted by Detructor View Post
        Need some reading comprehension there?
        He's talking about using the hardware encryption provided by this drivers code (the co-processor).

        Comment


        • #5
          Is this a separate chip, or is it something embedded in some class of CPUs or North/South-bridges, and we haven't had access to it until now?

          If it's already embedded, what chips, what vintages?

          Comment


          • #6
            i think it's for the future arm-on-opteron thing that amd was promoting a while ago

            Comment


            • #7
              Originally posted by movieman View Post
              But who's going to trust an encryption black box these days? Particularly from a US company?

              Even Intel's AES instructions are dubious; you can verify they encrypt properly in your tests, but you can't verify there isn't some hidden switch in a configuration register or undocumented instruction somewhere which will make them spew out plaintext instead.
              Stop being delusional. If you think the heads-of-state weren't in on the game to manage patterns from the NSA you truly are daft.

              Comment


              • #8
                AMD TrustZone

                This just sounds like the Linux drivers for the AMD TrustZone functionality that we've known about for some time.

                Comment


                • #9
                  Originally posted by movieman View Post
                  But who's going to trust an encryption black box these days? Particularly from a US company?
                  Idiots, that's who.

                  Comment


                  • #10
                    The only mentioned pci id, 0x1537, gives no results on the wonderland of google, nor in pciids.

                    So either a completely new block, or a well-hidden secret.

                    Comment

                    Working...
                    X