Mike Klein 8f506d3257 SSE filter speed improvements for bpp=3.
- memcpy-free implementations of load3() / store3().
    These should have less variance compiler to compiler.

    - call load3() only when needed at the end of a scanline.
    In the middle, we can use the faster load4(), ignoring that byte.
2016-04-04 16:10:09 -04:00
..

This "contrib" directory contains contributions which are not necessarily under
the libpng license, although all are open source.  They are not part of
libpng proper and are not used for building the library, although some are used
for testing the library via "make check".