On Sun, Oct 06, 2013 at 08:40:21PM +0200, Luca Barbato wrote: > On 06/10/13 19:28, Diego Biurrun wrote: > > x86_64 always has SSE2, so skip compiling in SIMD-optimized functions > > that will always be overridden by variants of the same function done > > with more advanced SIMD optimization types. > > Again, why didn't you change the logic of EXTERNAL_AMD3DNOW etc so you > spare lots of lines there?
I don't see how that could work at all for the assembly parts (which I overlooked in the first iteration of this patch). I also don't see how such a macro should figure out which of the parts it covers have already been set in another optimization block. Diego _______________________________________________ libav-devel mailing list libav-devel@libav.org https://lists.libav.org/mailman/listinfo/libav-devel