[FFmpeg-devel] avfilter/x86/vf_hflip : make macro and add AVX2

Henrik Gramner henrik at gramner.com
Wed Dec 13 18:18:08 EET 2017


On Wed, Dec 13, 2017 at 6:07 AM, Martin Vignali
<martin.vignali at gmail.com> wrote:
> +        vpermq  m1, [srcq + xq -     mmsize + %3], 0x4e; flip each lane at load
> +        vpermq  m2, [srcq + xq - 2 * mmsize + %3], 0x4e; flip each lane at load

Would doing 2x 128-bit movu + 2x vinserti128 be faster?


More information about the ffmpeg-devel mailing list