[Interest] SIMD accelerated wrappers for relevant Qt container classes (QVector, ...) or QByteArray?

Allan Sandfeld Jensen kde at carewolf.com
Wed May 10 12:21:48 CEST 2017


On Wednesday 10 May 2017, René J.V. Bertin wrote:
>  If memory serves me well the last time I looked at the intrinsic
> headerfiles shipped with clang on Mac they just defined macros expanding
> to inline assembly. So yeah, that should work.
> 
That is not how the simple intrinsics are defined any more. They instead use 
special SIMD extensions to tell the compiler to add two vectors of 8 integers 
together. When AVX2 is available that can be done with one instruction, but if 
-mavx2 is not used, then it will do it with SSE2 instead using 3 instructions 
or more.

`Allan



More information about the Interest mailing list