[x265] [PATCH] asm: assembly code for x265_pixel_satd_32x16

[yuvaraj at multicorewareinc.com]

# HG changeset patch
# User Yuvaraj Venkatesh <yuvaraj at multicorewareinc.com>
# Date 1384180564 -19800
#      Mon Nov 11 20:06:04 2013 +0530
# Node ID fffe6456b8a1a3b06685b1d755c98b3fe7005ab9
# Parent  1130addeb3b8f9daffa9c0f5d2852e1812169c02
asm: assembly code for x265_pixel_satd_32x16

diff -r 1130addeb3b8 -r fffe6456b8a1 source/common/x86/asm-primitives.cpp
--- a/source/common/x86/asm-primitives.cpp	Mon Nov 11 17:01:26 2013 +0530
+++ b/source/common/x86/asm-primitives.cpp	Mon Nov 11 20:06:04 2013 +0530
@@ -294,6 +294,7 @@
         HEVC_SATD(mmx2);
         p.satd[LUMA_12x16] = cmp<12, 16, 4, 16, x265_pixel_satd_4x16_mmx2>;
         p.satd[LUMA_32x8] = x265_pixel_satd_32x8_sse2;
+        p.satd[LUMA_32x16] = x265_pixel_satd_32x16_sse2;
         p.sa8d[BLOCK_4x4] = x265_pixel_satd_4x4_mmx2;
         p.frame_init_lowres_core = x265_frame_init_lowres_core_mmx2;
 
@@ -444,6 +445,7 @@
         p.satd[LUMA_4x16]   = x265_pixel_satd_4x16_sse4;
         p.satd[LUMA_12x16]  = cmp<12, 16, 4, 16, x265_pixel_satd_4x16_sse4>;
         p.satd[LUMA_32x8] = x265_pixel_satd_32x8_sse4;
+        p.satd[LUMA_32x16] = x265_pixel_satd_32x16_sse4;
         p.sa8d[BLOCK_8x8]   = x265_pixel_sa8d_8x8_sse4;
         p.sa8d[BLOCK_16x16] = x265_pixel_sa8d_16x16_sse4;
         SA8D_INTER_FROM_BLOCK(sse4);
@@ -469,6 +471,7 @@
         p.satd[LUMA_4x16]   = x265_pixel_satd_4x16_avx;
         p.satd[LUMA_12x16]  = cmp<12, 16, 4, 16, x265_pixel_satd_4x16_avx>;
         p.satd[LUMA_32x8] = x265_pixel_satd_32x8_avx;
+        p.satd[LUMA_32x16] = x265_pixel_satd_32x16_avx;
         p.sa8d[BLOCK_8x8]   = x265_pixel_sa8d_8x8_avx;
         p.sa8d[BLOCK_16x16] = x265_pixel_sa8d_16x16_avx;
         SA8D_INTER_FROM_BLOCK(avx);
diff -r 1130addeb3b8 -r fffe6456b8a1 source/common/x86/pixel-a.asm
--- a/source/common/x86/pixel-a.asm	Mon Nov 11 17:01:26 2013 +0530
+++ b/source/common/x86/pixel-a.asm	Mon Nov 11 20:06:04 2013 +0530
@@ -1761,7 +1761,70 @@
     call pixel_satd_16x4_internal
     SATD_END_SSE2 m10
 
+cglobal pixel_satd_32x16, 4,8,8    ;if WIN64 && notcpuflag(avx)
+    SATD_START_SSE2 m10, m7
+    mov r6, r0
+    mov r7, r2
+%if vertical
+    mova m7, [pw_00ff]
+%endif
+    call pixel_satd_16x4_internal
+    call pixel_satd_16x4_internal
+    call pixel_satd_16x4_internal
+    call pixel_satd_16x4_internal
+    lea r0, [r6 + 16]
+    lea r2, [r7 + 16]
+    call pixel_satd_16x4_internal
+    call pixel_satd_16x4_internal
+    call pixel_satd_16x4_internal
+    call pixel_satd_16x4_internal
+    SATD_END_SSE2 m10
+
 %else
+cglobal pixel_satd_32x16, 4,8,8
+%if WIN64                           ;if WIN64 && cpuflag(avx)
+    SATD_START_SSE2 m6, m7
+    mov r6, r0
+    mov r7, r2
+    call pixel_satd_8x8_internal
+    call pixel_satd_8x8_internal
+    lea r0, [r6 + 8]
+    lea r2, [r7 + 8]
+    call pixel_satd_8x8_internal
+    call pixel_satd_8x8_internal
+    lea r0, [r6 + 16]
+    lea r2, [r7 + 16]
+    call pixel_satd_8x8_internal
+    call pixel_satd_8x8_internal
+    lea r0, [r6 + 24]
+    lea r2, [r7 + 24]
+    call pixel_satd_8x8_internal
+    call pixel_satd_8x8_internal
+    SATD_END_SSE2 m6
+%else                               ;if !WIN64
+    SATD_START_SSE2 m6, m7
+    call pixel_satd_8x8_internal
+    call pixel_satd_8x8_internal
+    mov r0, r0mp
+    mov r2, r2mp
+    add r0, 8
+    add r2, 8
+    call pixel_satd_8x8_internal
+    call pixel_satd_8x8_internal
+    mov r0, r0mp
+    mov r2, r2mp
+    add r0, 16
+    add r2, 16
+    call pixel_satd_8x8_internal
+    call pixel_satd_8x8_internal
+    mov r0, r0mp
+    mov r2, r2mp
+    add r0, 24
+    add r2, 24
+    call pixel_satd_8x8_internal
+    call pixel_satd_8x8_internal
+    SATD_END_SSE2 m6
+%endif
 
 cglobal pixel_satd_32x8, 4,6,8    ;if !WIN64
     SATD_START_SSE2 m6, m7