commit 0768d98bdb473facab16e3c10967cdb5c356fb3a
author Deepa K G <deepa.kg@ittiam.com> Mon Feb 05 13:41:42 2018 +0530
committer Yunqing Wang <yunqingwang@google.com> Tue Feb 06 18:04:48 2018 +0000
tree ac8df0933863a457151c96b270b437cd903c94f6
parent 3ea816e8bb4ac1c0f236adbf9222a677b20ead51

Faster AVX2 implementation of motion compensation modules

Improvements have been made to av1_convolve_y_avx2 (~1.5x faster),
av1_convolve_y_sr_avx2 (~1.8x faster) and av1_convolve_2d_sr_avx2 (~1.3x faster).

Change-Id: Iaed764a7c4d069a4180c3edb0b1ac57ad36dad21

aom_dsp/aom_dsp.cmake

diff --git a/aom_dsp/aom_dsp.cmake b/aom_dsp/aom_dsp.cmake
index f61af74..6c1d89e 100644
--- a/aom_dsp/aom_dsp.cmake
+++ b/aom_dsp/aom_dsp.cmake
@@ -81,6 +81,7 @@
     "${AOM_ROOT}/aom_dsp/x86/intrapred_avx2.c"
     "${AOM_ROOT}/aom_dsp/x86/inv_txfm_avx2.c"
     "${AOM_ROOT}/aom_dsp/x86/common_avx2.h"
+    "${AOM_ROOT}/aom_dsp/x86/convolve_avx2.h"
     "${AOM_ROOT}/aom_dsp/x86/inv_txfm_common_avx2.h"
     "${AOM_ROOT}/aom_dsp/x86/txfm_common_avx2.h")