kfjahnke

What really surprised me that *division by one* should produce an erroneous result. This is the one case where multiplication with the reciprocal should be fine, and correct for all...

So my error was to assume that forming the reciprocal of 1.0f would yield 1.0f. 0.999756f explains the result: every multiplication produces a result which is just a tad below...

Hey, you still remember me! Yes, std::simd would be great, and I also need to be able to use clang++, it optimizes my code much better than g++.

> Sorry if you feel/felt ignored. I was occupied with stupid corner cases like from -ffast-math. wink I'm bad at working on more than one thing at a time... You...

Precisely. It's good for prototyping, and the 'prototype' may already be sufficiently performant. If a 'real' SIMD implementation arrives later on, it's easy to specialize to it, and then to...

std::array is a good candidate, but using a dedicated type like vspline::simd_type (which simply contains a 1D C-style array) is probably cleaner, removing a big fat type which may be...

> I was hoping to avoid the special vector types because they are not portable and don't work in MSVC, but that compiler anyway isn't known for autovectorization. > >...

> There appears to be a difference in goals here. We cannot have anything that risks a large speed reduction unless perhaps users opt-in to a slow mode, but there...

Am 23.08.21 um 14:02 schrieb Mingye Wang: > WRT writing things that compilers (usually) autovectorize, the > simd-everywhere/simde project > might be interesting to look at. Thanks for the hint!...

> We haven't needed div/mod yet. I'm curious as to what the use case is? Generic programming. If I don't have Div defined for all types, I need to special-case...