On Fri, Nov 04, 2016 at 10:41:57AM -0400, Lennart Sorensen wrote: > Well it's not the complete solution yet. > > I have it down to only deinterlace being miscompiled (it auto vectorized > Generic* functions with altivec instructions, which is not good). > > At the moment it looks like either the altivec stuff has to be #ifdef > out and compiled in a separeta pass like the yuv _altivec_ plugin, or > gcc #pragma has to be used to only enable altivec in the one function > (which I am trying now). #pragma seems a tad ugly though. > > I suspect gcc didn't always generate vector code automatically when > -maltivec was passed, but these days it does, even at -O2 it seems, > even though all the places I can find in the documentation says only -O3 > enables vectorization.
This would of course go much faster if I currently had access to a powerpc box rather than running in qemu. :) I miss the p710 at my previous job. -- Len Sorensen
