[x265] [PATCH] asm: avx2 code for sad[32x32] for 8bpp

[sumalatha at multicorewareinc.com]

# HG changeset patch
# User Sumalatha Polureddy<sumalatha at multicorewareinc.com>
# Date 1426482793 -19800
# Node ID 8b3ce08d1ba85d4d31fd48f52efba19c22462254
# Parent  6461985f33ac6fc5b205879bbb0f2a535226ca76
asm: avx2 code for sad[32x32] for 8bpp

sad[32x32]  44.69x   494.86          22114.45

diff -r 6461985f33ac -r 8b3ce08d1ba8 source/common/x86/asm-primitives.cpp
--- a/source/common/x86/asm-primitives.cpp	Sun Mar 15 11:58:32 2015 -0500
+++ b/source/common/x86/asm-primitives.cpp	Mon Mar 16 10:43:13 2015 +0530
@@ -1444,6 +1444,8 @@
         p.pu[LUMA_8x16].satd  = x265_pixel_satd_8x16_avx2;
         p.pu[LUMA_8x8].satd   = x265_pixel_satd_8x8_avx2;
 
+        p.pu[LUMA_32x32].sad = x265_pixel_sad_32x32_avx2;
+
         p.pu[LUMA_8x4].sad_x3 = x265_pixel_sad_x3_8x4_avx2;
         p.pu[LUMA_8x8].sad_x3 = x265_pixel_sad_x3_8x8_avx2;
         p.pu[LUMA_8x16].sad_x3 = x265_pixel_sad_x3_8x16_avx2;
diff -r 6461985f33ac -r 8b3ce08d1ba8 source/common/x86/sad-a.asm
--- a/source/common/x86/sad-a.asm	Sun Mar 15 11:58:32 2015 -0500
+++ b/source/common/x86/sad-a.asm	Mon Mar 16 10:43:13 2015 +0530
@@ -3897,5 +3897,34 @@
     movq            [r6 + 8], xm1
     RET
 
+INIT_YMM avx2
+cglobal pixel_sad_32x32, 4,5,5
+    xorps           m0, m0
+    mov             r4d, 16
+
+.loop
+    movu           m1, [r0]               ; row 0 of pix0
+    movu           m2, [r2]               ; row 0 of pix1
+    movu           m3, [r0 + r1]          ; row 1 of pix0
+    movu           m4, [r2 + r3]          ; row 1 of pix1
+
+    psadbw         m1, m2
+    psadbw         m3, m4
+    paddd          m0, m1
+    paddd          m0, m3
+
+    lea     r2,     [r2 + 2 * r3]
+    lea     r0,     [r0 + 2 * r1]
+
+    dec         r4d
+    jnz         .loop
+
+    vextracti128   xm1, m0, 1
+    paddd          xm0, xm1
+    pshufd         xm1, xm0, 2
+    paddd          xm0,xm1
+    movd            eax, xm0
+    RET
+
 
 %endif