on Mon Dec 24 2012, Mathias Gaunard <mathias.gaunard-AT-ens-lyon.org> wrote:
On 24/12/12 03:55, Dave Abrahams wrote:
depending on segment alignment
I give up. If you want to discuss this with me, please at least take the time to read what I said.
I did read it. Quite carefully, I thought. I probably used the wrong words. What I meant by "segment alignment" in this case was "whether or not the start and end fall into the same segment," which according to what you wrote, could amount to "whether or not there is more than a SIMD vector's worth of data."
It is probably simpler to drop the subject.
Probably, but neither of us will learn anything from this discussion if we do. I'm still eager to understand if, and how, my theory is full of hooey. Nobody else is stepping forward to explain it, so I guess you're the only guy who can.
Merry Christmas.
hohoho-ly y'rs, -- Dave Abrahams BoostPro Computing Software Development Training http://www.boostpro.com Clang/LLVM/EDG Compilers C++ Boost