reg_sad implementation using SSE2/SSE4.1 intrinsics

src/Makefile

@@ -5,7 +5,7 @@ ifeq (, $(ARCH))
 endif
 SYSTEM = $(shell uname -s)
 ASMFLAGS =
-DFLAGS = -O2 -g -Werror -DUSE_TILES=1
+DFLAGS = -O2 -g -Werror -DUSE_TILES=1 -march=native
 
 # ARCH related flags
 ifeq ($(ARCH), x86_64)

src/picture-sse2.c

@@ -0,0 +1,90 @@
+/*****************************************************************************
+ * This file is part of Kvazaar HEVC encoder.
+ *
+ * Copyright (C) 2013-2014 Tampere University of Technology and others (see
+ * COPYING file).
+ *
+ * Kvazaar is free software: you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as published
+ * by the Free Software Foundation.
+ *
+ * Kvazaar is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with Kvazaar.  If not, see <http://www.gnu.org/licenses/>.
+ ****************************************************************************/
+
+/*
+ * \file
+ */
+
+#include "picture.h"
+#include <immintrin.h>
+
+#ifdef __SSE2__
+static unsigned reg_sad(const pixel * const data1, const pixel * const data2,
+                        const int width, const int height, const unsigned stride1, const unsigned stride2)
+{
+  int y, x;
+  unsigned sad = 0;
+  __m128i sse_inc = _mm_setzero_si128 ();
+  long long int sse_inc_array[2];
+  
+  for (y = 0; y < height; ++y) {
+    for (x = 0; x <= width-16; x+=16) {
+      const __m128i a = _mm_loadu_si128((__m128i const*) &data1[y * stride1 + x]);
+      const __m128i b = _mm_loadu_si128((__m128i const*) &data2[y * stride2 + x]);
+      sse_inc = _mm_add_epi32(sse_inc, _mm_sad_epu8(a,b));
+    }
+    
+#ifdef __SSE4_1__
+    {
+      const __m128i a = _mm_loadu_si128((__m128i const*) &data1[y * stride1 + x]);
+      const __m128i b = _mm_loadu_si128((__m128i const*) &data2[y * stride2 + x]);
+      switch (((width - (width%2)) - x)/2) {
+        case 0:
+          break;
+        case 1:
+          sse_inc = _mm_add_epi32(sse_inc, _mm_sad_epu8(a, _mm_blend_epi16(a, b, 0x01)));
+          break;
+        case 2:
+          sse_inc = _mm_add_epi32(sse_inc, _mm_sad_epu8(a, _mm_blend_epi16(a, b, 0x03)));
+          break;
+        case 3:
+          sse_inc = _mm_add_epi32(sse_inc, _mm_sad_epu8(a, _mm_blend_epi16(a, b, 0x07)));
+          break;
+        case 4:
+          sse_inc = _mm_add_epi32(sse_inc, _mm_sad_epu8(a, _mm_blend_epi16(a, b, 0x0f)));
+          break;
+        case 5:
+          sse_inc = _mm_add_epi32(sse_inc, _mm_sad_epu8(a, _mm_blend_epi16(a, b, 0x1f)));
+          break;
+        case 6:
+          sse_inc = _mm_add_epi32(sse_inc, _mm_sad_epu8(a, _mm_blend_epi16(a, b, 0x3f)));
+          break;
+        case 7:
+          sse_inc = _mm_add_epi32(sse_inc, _mm_sad_epu8(a, _mm_blend_epi16(a, b, 0x7f)));
+          break;
+        default:
+          //Should not happen
+          assert(0);
+      }
+      x = (width - (width%2));
+    }
+#endif //__SSE4_1__
+
+    for (; x < width; ++x) {
+      sad += abs(data1[y * stride1 + x] - data2[y * stride2 + x]);
+    }
+  }
+  _mm_storeu_si128((__m128i*) sse_inc_array, sse_inc);
+  sad += sse_inc_array[0] + sse_inc_array[1];
+
+  return sad;
+}
+#else
+#error picture-sse2.c requires __SSE2__
+#endif //__SSE2__

src/picture.c

@@ -768,6 +768,13 @@ static unsigned hor_sad(const pixel *pic_data, const pixel *ref_data,
   return sad;
 }
 
+
+#if defined(__SSE2__)
+#include "picture-sse2.c"
+#elif defined(__ALTIVEC__)
+#include "picture-altivec.c"
+#else
+//Generic implementations
 /**
  * \brief Calculate Sum of Absolute Differences (SAD)
  *
@@ -782,20 +789,21 @@ static unsigned hor_sad(const pixel *pic_data, const pixel *ref_data,
  *
  * \returns Sum of Absolute Differences
  */
-static unsigned reg_sad(const pixel *data1, const pixel *data2,
-                        int width, int height, unsigned stride1, unsigned stride2)
+static unsigned reg_sad(const pixel * const data1, const pixel * const data2,
+                        const int width, const int height, const unsigned stride1, const unsigned stride2)
 {
   int y, x;
   unsigned sad = 0;
 
   for (y = 0; y < height; ++y) {
     for (x = 0; x < width; ++x) {
-      sad += abs((int)data1[y * stride1 + x] - (int)data2[y * stride2 + x]);
+      sad += abs(data1[y * stride1 + x] - data2[y * stride2 + x]);
     }
   }
 
   return sad;
 }
+#endif
 
 /**
  * \brief  Handle special cases of comparing blocks that are not completely