dimula73 at gmail.com
Mon Oct 29 12:18:39 UTC 2012
On Mon, Oct 29, 2012 at 3:13 PM, Andre Gemuend <scroogie at scroogie.de> wrote:
> and what about the idea that Sven posted: using the same integer
> optimizations in SIMD? Or does VC not provide 16bit unsigned integer?
> Because in that way it should automatically use the same optimizations in
> the fallback, shouldn't it?
Well, the integer solution might be quite good idea for some of the
composite ops. When I investigated into the topic, I thought about it, but
I faced with two problems I had no idea how to solve it:
1) The sse instruction set does not have integer division instruction. So
it will have to be workarounded somehow.
2) Integer multiplication instructions change the size of the operands. I
didn't know how to solve it properly.
I guess, both of these these problems are solvable, but right now I don't
know how to deal with them.
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the kimageshop