Quote:
Originally Posted by nitsuj
SSE optimized branch? Nice. Let me know if it's a bug in my code that you've spotted. I'd like to get bugs ironed out before I ReaPack it.
BTW, great job on JSFX. Making it fast enough to do this kind of processing is really cool.
|
Glad you like JSFX, thanks for making such great scripts!
The bug I saw is almost certainly on our side (if it works on the release but fails in my branch, then the problem is in my branch). The SSE implementation isn't anything fancy with SIMD, just about getting rid of old x87 code (which is especially important on x86_64 where the calling convention dictates xmm0 as return values, etc).
Also: I did improve the compile speed of ReEQ in the latest +dev pre-releases (on my i7 RMBP it was taking 1.2s to compile, which I reduced down to 120ms... which is still quite a bit longer than I'd like it to be.)