]> git.sesse.net Git - ffmpeg/commit
swscale/x86/output: add AVX2 version of yuv2nv12cX
authorNelson Gomez <nelson.gomez@microsoft.com>
Sun, 26 Apr 2020 02:37:02 +0000 (19:37 -0700)
committerJosh de Kock <josh@itanimul.li>
Sun, 14 Jun 2020 15:34:07 +0000 (16:34 +0100)
commitbc01337db4d196b2c3597bfd1c4431edb8779159
treed4f0c4f4a5065d8e86302fd668f805ea12d9ce24
parent7c39c3c1a6f35a6b47970417b1e273141eadb856
swscale/x86/output: add AVX2 version of yuv2nv12cX

256 bits is just wide enough to fit all the operands needed to vectorize
the software implementation, but AVX2 is needed to for a couple of
instructions like cross-lane permutation.

Output is bit-for-bit identical to C.

Signed-off-by: Nelson Gomez <nelson.gomez@microsoft.com>
libswscale/x86/output.asm
libswscale/x86/swscale.c