[FFmpeg-cvslog] x86/vf_v360: use a faster horizontal add in remap4_8bit_line_avx2

[James Almer]

ffmpeg | branch: master | James Almer <jamrial at gmail.com> | Fri Sep  6 11:18:45 2019 -0300| [4857688732e27e33fac58e2bd30989f20da2f181] | committer: James Almer

x86/vf_v360: use a faster horizontal add in remap4_8bit_line_avx2

Signed-off-by: James Almer <jamrial at gmail.com>

> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=4857688732e27e33fac58e2bd30989f20da2f181
---

 libavfilter/x86/vf_v360.asm | 11 ++++-------
 1 file changed, 4 insertions(+), 7 deletions(-)

diff --git a/libavfilter/x86/vf_v360.asm b/libavfilter/x86/vf_v360.asm
index f49702b603..a0936eb6dc 100644
--- a/libavfilter/x86/vf_v360.asm
+++ b/libavfilter/x86/vf_v360.asm
@@ -130,14 +130,11 @@ cglobal remap4_8bit_line, 7, 9, 11, dst, width, src, in_linesize, u, v, ker, x,
         pmulld          m4, m5
 
         paddd           m2, m4
-        vextracti128   xm1, m2, 1
-        paddd           m1, m2
-        phaddd          m1, m1
-        phaddd          m1, m1
-        psrld           m1, m1, 0xe
-        packuswb        m1, m1
+        HADDD           m2, m1
+        psrld           m2, m2, 0xe
+        packuswb        m2, m2
 
-        pextrb   [dstq+xq], xm1, 0
+        pextrb   [dstq+xq], xm2, 0
 
         add   xq, 1
         add   yq, 32