x86/mpegvideoencdsp: improve ff_pix_sum16_sse2
authorJames Almer <jamrial@gmail.com>
Wed, 1 Oct 2014 01:21:40 +0000 (22:21 -0300)
committerJames Almer <jamrial@gmail.com>
Wed, 1 Oct 2014 16:07:22 +0000 (13:07 -0300)
commitacebff8e5dc0789c228b10ffcae2f2eb6c30a91d
tree33bebe65a886d5ee908242045e51591d0f052229
parentf2e53808e329c403d1a409f3c35a93854064b6a2
x86/mpegvideoencdsp: improve ff_pix_sum16_sse2

~15% faster.

Also add an mmxext version that takes advantage of the new code, and
build it alongside with the mmx version only on x86_32.

Reviewed-by: Michael Niedermayer <michaelni@gmx.at>
Signed-off-by: James Almer <jamrial@gmail.com>
libavcodec/x86/mpegvideoencdsp.asm
libavcodec/x86/mpegvideoencdsp_init.c