1. 12
  1.  

  2. 4

    The LKML thread on this patchset is miserably long. Linus seems to be super upset about Secure Boot, and I can’t quite understand why.

    https://www.spinics.net/lists/kernel/msg2766909.html is a nice succinct message from Kees Cook explaining why this is useful.

    1. 2

      Linus seems to be super upset about Secure Boot, and I can’t quite understand why.

      My understanding was that he isn’t upset about secure boot itself, but rather why the patchset was necessary, and Garrett’s response ended up being something along the lines of “why not” and “just disable it if you don’t like it” rather than actual arguments to back it up. In his reply to Linus’s outburst yesterday, Garrett ended up accusing Linus of not accepting the patch due to “political” reasons instead of technical ones.

      Like you said, the thread is miserably long and difficult to keep up with if you’re not fully aware of the technical details of what they’re talking about (like me). I’m certain I’ve missed some finer details on this subject, and I definitely don’t know the technical aspects of what’s going on - I’m just reporting what frustrations I’ve seen from Linus.

      1. 3

        I found mjg’s explanation pretty clear:

        1. SecureBoot without kernel lockdown has glaring security holes, so we should plug them
        2. kernel lockdown without a secured bootchain has an easy bypass, so we shouldn’t enable it and give people a false sense of security
        3. Enabling both is good.
        4. Enabling zero is silly, but has very clear security properties that won’t confuse anyone.