tree 505ebf23492a78c8780e297711e0d07f926229e2
parent 9424338d4d9388074612b909b50c72d3737ebb65
author Aniket Dhok <aniket.dhok@ittiam.com> 1556164768 +0530
committer Yunqing Wang <yunqingwang@google.com> 1556840161 +0000

Added AVX2 variant for av1_warp_affine_sse4_1

For speed = 1, 2, 3 and 4 presets observed encode time reduction of
0.8%, 0.95%, 0.2% and 0.2% (averaged across multiple test cases).

Module gains improved by factor of ~1.4x w.r.t SSE4_1 module

Change-Id: I5007f2a96347844e969a5cdcf9c7e3a2928bf8f3