[x265] [PATCH Review only] asm: pixel_add_pp routine for 8x6 block size

[murugan at multicorewareinc.com]

# HG changeset patch
# User Murugan Vairavel <murugan at multicorewareinc.com>
# Date 1383653386 -19800
#      Tue Nov 05 17:39:46 2013 +0530
# Node ID 264f215c3503097b8ec4cebfaade85caaa540048
# Parent  1f3de94bee30eebab7f10df37e674fa7360803b1
asm: pixel_add_pp routine for 8x6 block size

diff -r 1f3de94bee30 -r 264f215c3503 source/common/x86/pixel-add8.asm
--- a/source/common/x86/pixel-add8.asm	Tue Nov 05 17:36:58 2013 +0530
+++ b/source/common/x86/pixel-add8.asm	Tue Nov 05 17:39:46 2013 +0530
@@ -90,3 +90,40 @@
 
 RET
 
+;-----------------------------------------------------------------------------
+; void pixel_add_pp_c_8x6(pixel *dest, intptr_t destride, pixel *src0, pixel *src1, intptr_t srcstride0, intptr_t srcstride1);
+;-----------------------------------------------------------------------------
+%macro PIXELADD_PP_W8_H2 2
+INIT_XMM sse2
+cglobal pixel_add_pp_%1x%2, 4, 7, 4, dest, deststride, src0, src1
+
+mov    r4d,    r4m
+mov    r5d,    r5m
+mov    r6d,    %2
+
+.loop
+
+    movh       m0,    [r2]
+    movh       m1,    [r3]
+
+    movh       m2,    [r2 + r4]
+    movh       m3,    [r3 + r5]
+
+    paddusb    m0,    m1
+    paddusb    m2,    m3
+
+    movh    [r0],         m0
+    movh    [r0 + r1],    m2
+
+    lea     r2,           [r2 + 2 * r4]
+    lea     r3,           [r3 + 2 * r5]
+    lea     r0,           [r0 + 2 * r1]
+
+    sub      r6d,          2
+
+jnz    .loop
+
+RET
+%endmacro
+
+PIXELADD_PP_W8_H2 8, 6