On Sat, Jan 14, 2012 at 1:46 PM, Eric Dumazet <eric.dumazet@xxxxxxxxx> wrote:
>
> This is too risky, and we provided an alternate patch, not just for fun.
Did you see the second patch?
The one that got rid of the *stupid* 80-entry array?
I don't know why so many sha implementations do that idiotic full
array, when the circular one is much better.
In fact, the 16-entry circular array allows machines with lots of
registers to keep all the state in registers and the C implementation
can often be as good as hand-tuned assembly. At least that's true for
sha1, I'm not sure you can do the same with sha512.
But that actually *requires* that the 16-entry array be done on the
stack as an automatic array. Anything else, and the compiler won't be
able to do it.
Linus
--
To unsubscribe from this list: send the line "unsubscribe linux-crypto" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
