On Thu, Apr 30, 2009 at 5:24 PM, Ariane van der Steldt <ari...@stack.nl> wrote: > I think if you are going to differentiate on that level, you'll not get > it in the kernel.
That's fine. This is a personal project primarily intended for my own benefit in learning more about hacking on the kernel. If it actually improves anything, that's just a bonus. :) > Differentiating on that level would mean we have an > enormous job to keep all different implementations working. Well, fortunately the majority of the code has a fixed target: look at a key, look at some input bytes, write some output bytes. The definition of AES or how fast instructions are on the Core 2 aren't going to change. :) > Keep in mind > that OpenBSD is not an intel-central world: we also have sparc64, sparc, > alpha, powerpc etc. Sure, and there are optimized AES implementations for those too. My immediate goal of just focusing on i386/amd64 is just because I have that hardware available. > The difference in speed is likely to only matter a number of nanoseconds > and I think you'll really notice that that speed-up is dwarfed by other > possible optimizations, ones that also affect other platforms positively. I guess we'll see. :) Like I said, just a personal project.
