Merge "First partial snapshot of vectorized 8-tap filtering." into experimental

4 files changed, 467 insertions, 45 deletions

diff --git a/vp8/common/filter.c b/vp8/common/filter.c
index 8b96c82a2..0bc88e5dd 100644
--- a/vp8/common/filter.c
+++ b/vp8/common/filter.c
@@ -12,6 +12,7 @@
 #include <stdlib.h>
 #include "filter.h"
 #include "vpx_ports/mem.h"
+#include "vpx_rtcd.h"
 
 DECLARE_ALIGNED(16, const short, vp8_bilinear_filters[SUBPEL_SHIFTS][2]) = {
   { 128,   0 },
@@ -511,13 +512,10 @@ static const unsigned int filter_max_width = 16;
 
 static void vp8_filter_block2d_8_c
 (
- const unsigned char *src_ptr,
- const unsigned int   src_stride,
- const short *HFilter,
- const short *VFilter,
+ const unsigned char *src_ptr, const unsigned int   src_stride,
+ const short *HFilter, const short *VFilter,
  const filter_size_t filter_size,
- unsigned char *dst_ptr,
- unsigned int   dst_stride
+ unsigned char *dst_ptr, unsigned int   dst_stride
 ) {
   const unsigned int output_width = filter_size_to_wh[filter_size][0];
   const unsigned int output_height = filter_size_to_wh[filter_size][1];
@@ -609,6 +607,50 @@ static void vp8_filter_block2d_8_c
   }
 }
 
+void vp8_filter_block2d_4x4_8_c
+(
+ const unsigned char *src_ptr, const unsigned int src_stride,
+ const short *HFilter_aligned16, const short *VFilter_aligned16,
+ unsigned char *dst_ptr, unsigned int dst_stride
+) {
+  vp8_filter_block2d_8_c(src_ptr, src_stride,
+                         HFilter_aligned16, VFilter_aligned16,
+                         VPX_FILTER_4x4, dst_ptr, dst_stride);
+}
+
+void vp8_filter_block2d_8x4_8_c
+(
+ const unsigned char *src_ptr, const unsigned int src_stride,
+ const short *HFilter_aligned16, const short *VFilter_aligned16,
+ unsigned char *dst_ptr, unsigned int dst_stride
+) {
+  vp8_filter_block2d_8_c(src_ptr, src_stride,
+                         HFilter_aligned16, VFilter_aligned16,
+                         VPX_FILTER_8x4, dst_ptr, dst_stride);
+}
+
+void vp8_filter_block2d_8x8_8_c
+(
+ const unsigned char *src_ptr, const unsigned int src_stride,
+ const short *HFilter_aligned16, const short *VFilter_aligned16,
+ unsigned char *dst_ptr, unsigned int dst_stride
+) {
+  vp8_filter_block2d_8_c(src_ptr, src_stride,
+                         HFilter_aligned16, VFilter_aligned16,
+                         VPX_FILTER_8x8, dst_ptr, dst_stride);
+}
+
+void vp8_filter_block2d_16x16_8_c
+(
+ const unsigned char *src_ptr, const unsigned int src_stride,
+ const short *HFilter_aligned16, const short *VFilter_aligned16,
+ unsigned char *dst_ptr, unsigned int dst_stride
+) {
+  vp8_filter_block2d_8_c(src_ptr, src_stride,
+                         HFilter_aligned16, VFilter_aligned16,
+                         VPX_FILTER_16x16, dst_ptr, dst_stride);
+}
+
 static void vp8_block2d_average_c
 (
   unsigned char *src,
@@ -629,7 +671,6 @@ static void vp8_block2d_average_c
   }
 }
 
-#define vp8_filter_block2d_8 vp8_filter_block2d_8_c
 #define vp8_block2d_average vp8_block2d_average_c
 
 void vp8_eighttap_predict_c
@@ -647,9 +688,9 @@ void vp8_eighttap_predict_c
   HFilter = vp8_sub_pel_filters_8[xoffset];
   VFilter = vp8_sub_pel_filters_8[yoffset];
 
-  vp8_filter_block2d_8(src_ptr, src_pixels_per_line,
-                       HFilter, VFilter, VPX_FILTER_4x4,
-                       dst_ptr, dst_pitch);
+  vp8_filter_block2d_4x4_8(src_ptr, src_pixels_per_line,
+                           HFilter, VFilter,
+                           dst_ptr, dst_pitch);
 }
 
 void vp8_eighttap_predict_avg4x4_c
@@ -665,9 +706,9 @@ void vp8_eighttap_predict_avg4x4_c
   const short  *VFilter = vp8_sub_pel_filters_8[yoffset];
   unsigned char tmp[4 * 4];
 
-  vp8_filter_block2d_8(src_ptr, src_pixels_per_line,
-                       HFilter, VFilter, VPX_FILTER_4x4,
-                       tmp, 4);
+  vp8_filter_block2d_4x4_8(src_ptr, src_pixels_per_line,
+                           HFilter, VFilter,
+                           tmp, 4);
   vp8_block2d_average(tmp, 4, dst_ptr, dst_pitch, VPX_FILTER_4x4);
 }
 
@@ -686,9 +727,9 @@ void vp8_eighttap_predict_sharp_c
   HFilter = vp8_sub_pel_filters_8s[xoffset];
   VFilter = vp8_sub_pel_filters_8s[yoffset];
 
-  vp8_filter_block2d_8(src_ptr, src_pixels_per_line,
-                       HFilter, VFilter, VPX_FILTER_4x4,
-                       dst_ptr, dst_pitch);
+  vp8_filter_block2d_4x4_8(src_ptr, src_pixels_per_line,
+                           HFilter, VFilter,
+                           dst_ptr, dst_pitch);
 }
 
 void vp8_eighttap_predict_avg4x4_sharp_c
@@ -704,9 +745,9 @@ void vp8_eighttap_predict_avg4x4_sharp_c
   const short  *VFilter = vp8_sub_pel_filters_8s[yoffset];
   unsigned char tmp[4 * 4];
 
-  vp8_filter_block2d_8(src_ptr, src_pixels_per_line,
-                       HFilter, VFilter, VPX_FILTER_4x4,
-                       tmp, 4);
+  vp8_filter_block2d_4x4_8(src_ptr, src_pixels_per_line,
+                           HFilter, VFilter,
+                           tmp, 4);
   vp8_block2d_average(tmp, 4, dst_ptr, dst_pitch, VPX_FILTER_4x4);
 }
 
@@ -722,9 +763,9 @@ void vp8_eighttap_predict8x8_c
   const short  *HFilter = vp8_sub_pel_filters_8[xoffset];
   const short  *VFilter = vp8_sub_pel_filters_8[yoffset];
 
-  vp8_filter_block2d_8(src_ptr, src_pixels_per_line,
-                       HFilter, VFilter, VPX_FILTER_8x8,
-                       dst_ptr, dst_pitch);
+  vp8_filter_block2d_8x8_8(src_ptr, src_pixels_per_line,
+                           HFilter, VFilter,
+                           dst_ptr, dst_pitch);
 }
 
 void vp8_eighttap_predict8x8_sharp_c
@@ -739,9 +780,9 @@ void vp8_eighttap_predict8x8_sharp_c
   const short  *HFilter = vp8_sub_pel_filters_8s[xoffset];
   const short  *VFilter = vp8_sub_pel_filters_8s[yoffset];
 
-  vp8_filter_block2d_8(src_ptr, src_pixels_per_line,
-                       HFilter, VFilter, VPX_FILTER_8x8,
-                       dst_ptr, dst_pitch);
+  vp8_filter_block2d_8x8_8(src_ptr, src_pixels_per_line,
+                           HFilter, VFilter,
+                           dst_ptr, dst_pitch);
 }
 
 void vp8_eighttap_predict_avg8x8_c
@@ -757,9 +798,9 @@ void vp8_eighttap_predict_avg8x8_c
   const short  *HFilter = vp8_sub_pel_filters_8[xoffset];
   const short  *VFilter = vp8_sub_pel_filters_8[yoffset];
 
-  vp8_filter_block2d_8(src_ptr, src_pixels_per_line,
-                       HFilter, VFilter, VPX_FILTER_8x8,
-                       tmp, 8);
+  vp8_filter_block2d_8x8_8(src_ptr, src_pixels_per_line,
+                           HFilter, VFilter,
+                           tmp, 8);
   vp8_block2d_average(tmp, 8, dst_ptr, dst_pitch, VPX_FILTER_8x8);
 }
 
@@ -776,9 +817,9 @@ void vp8_eighttap_predict_avg8x8_sharp_c
   const short  *HFilter = vp8_sub_pel_filters_8s[xoffset];
   const short  *VFilter = vp8_sub_pel_filters_8s[yoffset];
 
-  vp8_filter_block2d_8(src_ptr, src_pixels_per_line,
-                       HFilter, VFilter, VPX_FILTER_8x8,
-                       tmp, 8);
+  vp8_filter_block2d_8x8_8(src_ptr, src_pixels_per_line,
+                           HFilter, VFilter,
+                           tmp, 8);
   vp8_block2d_average(tmp, 8, dst_ptr, dst_pitch, VPX_FILTER_8x8);
 }
 
@@ -795,9 +836,9 @@ void vp8_eighttap_predict8x4_c
   const short  *HFilter = vp8_sub_pel_filters_8[xoffset];
   const short  *VFilter = vp8_sub_pel_filters_8[yoffset];
 
-  vp8_filter_block2d_8(src_ptr, src_pixels_per_line,
-                       HFilter, VFilter, VPX_FILTER_8x4,
-                       dst_ptr, dst_pitch);
+  vp8_filter_block2d_8x4_8(src_ptr, src_pixels_per_line,
+                           HFilter, VFilter,
+                           dst_ptr, dst_pitch);
 }
 
 void vp8_eighttap_predict8x4_sharp_c
@@ -812,9 +853,9 @@ void vp8_eighttap_predict8x4_sharp_c
   const short  *HFilter = vp8_sub_pel_filters_8s[xoffset];
   const short  *VFilter = vp8_sub_pel_filters_8s[yoffset];
 
-  vp8_filter_block2d_8(src_ptr, src_pixels_per_line,
-                       HFilter, VFilter, VPX_FILTER_8x4,
-                       dst_ptr, dst_pitch);
+  vp8_filter_block2d_8x4_8(src_ptr, src_pixels_per_line,
+                           HFilter, VFilter,
+                           dst_ptr, dst_pitch);
 }
 
 void vp8_eighttap_predict16x16_c
@@ -829,8 +870,8 @@ void vp8_eighttap_predict16x16_c
   const short  *HFilter = vp8_sub_pel_filters_8[xoffset];
   const short  *VFilter = vp8_sub_pel_filters_8[yoffset];
 
-  vp8_filter_block2d_8(src_ptr, src_pixels_per_line,
-                       HFilter, VFilter, VPX_FILTER_16x16,
+  vp8_filter_block2d_16x16_8(src_ptr, src_pixels_per_line,
+                       HFilter, VFilter,
                        dst_ptr, dst_pitch);
 }
 
@@ -846,8 +887,8 @@ void vp8_eighttap_predict16x16_sharp_c
   const short  *HFilter = vp8_sub_pel_filters_8s[xoffset];
   const short  *VFilter = vp8_sub_pel_filters_8s[yoffset];
 
-  vp8_filter_block2d_8(src_ptr, src_pixels_per_line,
-                       HFilter, VFilter, VPX_FILTER_16x16,
+  vp8_filter_block2d_16x16_8(src_ptr, src_pixels_per_line,
+                       HFilter, VFilter,
                        dst_ptr, dst_pitch);
 }
 
@@ -865,8 +906,8 @@ void vp8_eighttap_predict_avg16x16_c
   const short  *HFilter = vp8_sub_pel_filters_8[xoffset];
   const short  *VFilter = vp8_sub_pel_filters_8[yoffset];
 
-  vp8_filter_block2d_8(src_ptr, src_pixels_per_line,
-                       HFilter, VFilter, VPX_FILTER_16x16,
+  vp8_filter_block2d_16x16_8(src_ptr, src_pixels_per_line,
+                       HFilter, VFilter,
                        tmp, 16);
   vp8_block2d_average(tmp, 16, dst_ptr, dst_pitch, VPX_FILTER_16x16);
 }
@@ -884,8 +925,8 @@ void vp8_eighttap_predict_avg16x16_sharp_c
   const short  *HFilter = vp8_sub_pel_filters_8s[xoffset];
   const short  *VFilter = vp8_sub_pel_filters_8s[yoffset];
 
-  vp8_filter_block2d_8(src_ptr, src_pixels_per_line,
-                       HFilter, VFilter, VPX_FILTER_16x16,
+  vp8_filter_block2d_16x16_8(src_ptr, src_pixels_per_line,
+                       HFilter, VFilter,
                        tmp, 16);
   vp8_block2d_average(tmp, 16, dst_ptr, dst_pitch, VPX_FILTER_16x16);
 }
diff --git a/vp8/common/rtcd_defs.sh b/vp8/common/rtcd_defs.sh
index 95148c86f..1cb5de311 100644
--- a/vp8/common/rtcd_defs.sh
+++ b/vp8/common/rtcd_defs.sh
@@ -4,3 +4,18 @@ struct blockd;
 EOF
 }
 forward_decls common_forward_decls
+
+prototype void vp8_filter_block2d_4x4_8 "const unsigned char *src_ptr, const unsigned int src_stride, const short *HFilter_aligned16, const short *VFilter_aligned16, unsigned char *dst_ptr, unsigned int dst_stride"
+prototype void vp8_filter_block2d_8x4_8 "const unsigned char *src_ptr, const unsigned int src_stride, const short *HFilter_aligned16, const short *VFilter_aligned16, unsigned char *dst_ptr, unsigned int dst_stride"
+prototype void vp8_filter_block2d_8x8_8 "const unsigned char *src_ptr, const unsigned int src_stride, const short *HFilter_aligned16, const short *VFilter_aligned16, unsigned char *dst_ptr, unsigned int dst_stride"
+prototype void vp8_filter_block2d_16x16_8 "const unsigned char *src_ptr, const unsigned int src_stride, const short *HFilter_aligned16, const short *VFilter_aligned16, unsigned char *dst_ptr, unsigned int dst_stride"
+
+# At the very least, MSVC 2008 has compiler bug exhibited by this code; code
+# compiles warning free but a dissassembly of generated code show bugs. To be
+# on the safe side, only enabled when compiled with 'gcc'.
+if [ "$CONFIG_GCC" = "yes" ]; then
+    specialize vp8_filter_block2d_4x4_8 sse4_1
+    specialize vp8_filter_block2d_8x4_8 sse4_1
+    specialize vp8_filter_block2d_8x8_8 sse4_1
+    specialize vp8_filter_block2d_16x16_8 sse4_1
+fi
diff --git a/vp8/common/x86/filter_sse4.c b/vp8/common/x86/filter_sse4.c
new file mode 100644
index 000000000..40e37d69c
--- /dev/null
+++ b/vp8/common/x86/filter_sse4.c
@@ -0,0 +1,361 @@
+/*
+ *  Copyright (c) 2012 The WebM project authors. All Rights Reserved.
+ *
+ *  Use of this source code is governed by a BSD-style license
+ *  that can be found in the LICENSE file in the root of the source
+ *  tree. An additional intellectual property rights grant can be found
+ *  in the file PATENTS.  All contributing project authors may
+ *  be found in the AUTHORS file in the root of the source tree.
+ */
+
+#include <assert.h> // for alignment checks
+#include <smmintrin.h> // SSE4.1
+#include "vp8/common/filter.h"
+#include "vpx_ports/mem.h" // for DECLARE_ALIGNED
+#include "vpx_rtcd.h"
+
+// TODO(cd): After cleanup, commit faster versions for non 4x4 size. This is
+//           just a quick partial snapshot so that other can already use some
+//           speedup.
+// TODO(cd): Use vectorized 8 tap filtering code as speedup to pure C 6 tap
+//           filtering.
+// TODO(cd): Reduce source size by using macros instead of current code
+//           duplication.
+// TODO(cd): Add some comments, better variable naming.
+// TODO(cd): Maybe use _mm_maddubs_epi16 if smaller filter coeficients (no sum
+//           of positive above 128), or have higher precision filter
+//           coefficients.
+// TODO(cd): Remove use of _mm_extract_epi32 and _mm_extract_epi64, to not
+//           require SSE4.1
+// TODO(cd): Remove use of _mm_shuffle_epi8 to not require SSSE3
+
+DECLARE_ALIGNED(16, static const unsigned char, mask0123_c[16]) = {
+  0x00, 0x01,
+  0x01, 0x02,
+  0x02, 0x03,
+  0x03, 0x04,
+  0x02, 0x03,
+  0x03, 0x04,
+  0x04, 0x05,
+  0x05, 0x06,
+};
+DECLARE_ALIGNED(16, static const unsigned char, mask4567_c[16]) = {
+  0x04, 0x05,
+  0x05, 0x06,
+  0x06, 0x07,
+  0x07, 0x08,
+  0x06, 0x07,
+  0x07, 0x08,
+  0x08, 0x09,
+  0x09, 0x0A,
+};
+DECLARE_ALIGNED(16, static const unsigned int, rounding_c[4]) = {
+  VP8_FILTER_WEIGHT >> 1,
+  VP8_FILTER_WEIGHT >> 1,
+  VP8_FILTER_WEIGHT >> 1,
+  VP8_FILTER_WEIGHT >> 1,
+};
+
+void vp8_filter_block2d_4x4_8_sse4_1
+(
+ const unsigned char *src_ptr, const unsigned int src_stride,
+ const short *HFilter_aligned16, const short *VFilter_aligned16,
+ unsigned char *dst_ptr, unsigned int dst_stride
+) {
+  const unsigned int output_height_div4 = 1;
+
+  DECLARE_ALIGNED(16, unsigned char, intermediate_buffer[4 * 12]);
+  const int kInterp_Extend = 4;
+  const unsigned int output_height = output_height_div4 * 4 - 1 +
+                                     2 * kInterp_Extend;
+
+  const __m128i zero = _mm_set1_epi16(0);
+  const __m128i mask0123 = _mm_load_si128((const __m128i *)mask0123_c);
+  const __m128i mask4567 = _mm_load_si128((const __m128i *)mask4567_c);
+  const __m128i rounding = _mm_load_si128((const __m128i *)rounding_c);
+
+  unsigned int i;
+  // check alignment
+  assert(0 == ((long)HFilter_aligned16)%16);
+  assert(0 == ((long)VFilter_aligned16)%16);
+
+  {
+    __m128i transpose3_0;
+    __m128i transpose3_1;
+    __m128i transpose3_2;
+    __m128i transpose3_3;
+
+    // Horizontal pass (src -> intermediate).
+    {
+      unsigned char *output_ptr = intermediate_buffer;
+      const __m128i HFilter = _mm_load_si128((const __m128i *)HFilter_aligned16);
+      // get first two columns filter coefficients
+      __m128i fil01 = _mm_shuffle_epi32(HFilter, _MM_SHUFFLE(0, 0, 0, 0));
+      __m128i fil23 = _mm_shuffle_epi32(HFilter, _MM_SHUFFLE(1, 1, 1, 1));
+      __m128i fil45 = _mm_shuffle_epi32(HFilter, _MM_SHUFFLE(2, 2, 2, 2));
+      __m128i fil67 = _mm_shuffle_epi32(HFilter, _MM_SHUFFLE(3, 3, 3, 3));
+      src_ptr -= (kInterp_Extend - 1) * src_stride + (kInterp_Extend - 1);
+
+      for (i = 0; i < output_height; i++) {
+        //load pixels
+        __m128i src  = _mm_loadu_si128((const __m128i *)src_ptr);
+        // extract the ones used for first column
+        __m128i src0123 = _mm_shuffle_epi8(src, mask0123);
+        __m128i src4567 = _mm_shuffle_epi8(src, mask4567);
+        __m128i src01_16 = _mm_unpacklo_epi8(src0123, zero);
+        __m128i src23_16 = _mm_unpackhi_epi8(src0123, zero);
+        __m128i src45_16 = _mm_unpacklo_epi8(src4567, zero);
+        __m128i src67_16 = _mm_unpackhi_epi8(src4567, zero);
+        // multiply accumulate them
+        __m128i mad01 = _mm_madd_epi16(src01_16, fil01);
+        __m128i mad23 = _mm_madd_epi16(src23_16, fil23);
+        __m128i mad45 = _mm_madd_epi16(src45_16, fil45);
+        __m128i mad67 = _mm_madd_epi16(src67_16, fil67);
+        __m128i mad0123 = _mm_add_epi32(mad01, mad23);
+        __m128i mad4567 = _mm_add_epi32(mad45, mad67);
+        __m128i mad_all = _mm_add_epi32(mad0123, mad4567);
+        mad_all = _mm_add_epi32(mad_all, rounding);
+        mad_all = _mm_srai_epi32(mad_all, VP8_FILTER_SHIFT);
+        mad_all = _mm_packs_epi32(mad_all, mad_all);
+        mad_all = _mm_packus_epi16(mad_all, mad_all);
+        *((unsigned int *)output_ptr) = _mm_extract_epi32(mad_all, 0);
+        //TODO(cd): look into Ronald's comment:
+        //    future suggestion: use movd, not pextrd(0).
+        //
+        //    Alternatively, you could unroll this loop somewhat to handle 2
+        //    or 4 lines at a time, so that the packs_epi32() and the
+        //    packus_epi16() handle a full register worth of data. Then again,
+        //    you might have to specialcase the last line since you have 11
+        //    lines to handle here, and you don't want to handle a 12th dummy
+        //    line, so overall I'm not sure it's worth it. Use your best
+        //    judgement. :-).
+
+        // next row
+        src_ptr += src_stride;
+        output_ptr += 4;
+      }
+    }
+
+    // Transpose result (intermediate -> transpose3_x)
+    {
+      const __m128i srcA  = _mm_load_si128((__m128i *)(&intermediate_buffer[ 0]));
+      const __m128i srcB  = _mm_load_si128((__m128i *)(&intermediate_buffer[16]));
+      const __m128i srcC  = _mm_load_si128((__m128i *)(&intermediate_buffer[32]));
+      // 00 01 02 03 10 11 12 13 20 21 22 23 30 31 32 33
+      // 40 41 42 43 50 51 52 53 60 61 62 63 70 71 72 73
+      // 80 81 82 83 90 91 92 93 A0 A1 A2 A3 xx xx xx xx
+      const __m128i transpose0_0 = _mm_unpacklo_epi8(srcA, srcB);
+      const __m128i transpose0_1 = _mm_unpackhi_epi8(srcA, srcB);
+      const __m128i transpose0_2 = _mm_unpacklo_epi8(srcC, srcC);
+      const __m128i transpose0_3 = _mm_unpackhi_epi8(srcC, srcC);
+      // 00 40 01 41 02 42 03 43 10 50 11 51 12 52 13 53
+      // 20 60 21 61 22 62 23 63 30 70 31 71 32 72 33 73
+      // 80 xx 81 xx 82 xx 83 xx 90 xx 91 xx 92 xx 93 xx
+      // A0 xx A1 xx A2 xx A3 xx xx xx xx xx xx xx xx xx
+      const __m128i transpose1_0 = _mm_unpacklo_epi8(transpose0_0, transpose0_1);
+      const __m128i transpose1_1 = _mm_unpackhi_epi8(transpose0_0, transpose0_1);
+      const __m128i transpose1_2 = _mm_unpacklo_epi8(transpose0_2, transpose0_3);
+      const __m128i transpose1_3 = _mm_unpackhi_epi8(transpose0_2, transpose0_3);
+      // 00 20 40 60 01 21 41 61 02 22 42 62 03 23 43 63
+      // 10 30 50 70 11 31 51 71 12 32 52 72 13 33 53 73
+      // 80 A0 xx xx 81 A1 xx xx 82 A2 xx xx 83 A3 xx xx
+      // 90 xx xx xx 91 xx xx xx 92 xx xx xx 93 xx xx xx
+      const __m128i transpose2_0 = _mm_unpacklo_epi8(transpose1_0, transpose1_1);
+      const __m128i transpose2_1 = _mm_unpackhi_epi8(transpose1_0, transpose1_1);
+      const __m128i transpose2_2 = _mm_unpacklo_epi8(transpose1_2, transpose1_3);
+      const __m128i transpose2_3 = _mm_unpackhi_epi8(transpose1_2, transpose1_3);
+      // 00 10 20 30 40 50 60 70 01 11 21 31 41 51 61 71
+      // 02 12 22 32 42 52 62 72 03 13 23 33 43 53 63 73
+      // 80 90 A0 xx xx xx xx xx 81 91 A1 xx xx xx xx xx
+      // 82 92 A2 xx xx xx xx xx 83 93 A3 xx xx xx xx xx
+      transpose3_0 = _mm_castps_si128(
+                            _mm_shuffle_ps(_mm_castsi128_ps(transpose2_0),
+                                           _mm_castsi128_ps(transpose2_2),
+                                           _MM_SHUFFLE(1, 0, 1, 0)));
+      transpose3_1 = _mm_castps_si128(
+                            _mm_shuffle_ps(_mm_castsi128_ps(transpose2_0),
+                                           _mm_castsi128_ps(transpose2_2),
+                                           _MM_SHUFFLE(3, 2, 3, 2)));
+      transpose3_2 = _mm_castps_si128(
+                            _mm_shuffle_ps(_mm_castsi128_ps(transpose2_1),
+                                           _mm_castsi128_ps(transpose2_3),
+                                           _MM_SHUFFLE(1, 0, 1, 0)));
+      transpose3_3 = _mm_castps_si128(
+                            _mm_shuffle_ps(_mm_castsi128_ps(transpose2_1),
+                                           _mm_castsi128_ps(transpose2_3),
+                                           _MM_SHUFFLE(3, 2, 3, 2)));
+      // 00 10 20 30 40 50 60 70 80 90 A0 xx xx xx xx xx
+      // 01 11 21 31 41 51 61 71 81 91 A1 xx xx xx xx xx
+      // 02 12 22 32 42 52 62 72 82 92 A2 xx xx xx xx xx
+      // 03 13 23 33 43 53 63 73 83 93 A3 xx xx xx xx xx
+    }
+
+    // Vertical pass (transpose3_x -> dst).
+    {
+      const __m128i VFilter = _mm_load_si128((const __m128i *)VFilter_aligned16);
+      // get first two columns filter coefficients
+      __m128i fil01 = _mm_shuffle_epi32(VFilter, _MM_SHUFFLE(0, 0, 0, 0));
+      __m128i fil23 = _mm_shuffle_epi32(VFilter, _MM_SHUFFLE(1, 1, 1, 1));
+      __m128i fil45 = _mm_shuffle_epi32(VFilter, _MM_SHUFFLE(2, 2, 2, 2));
+      __m128i fil67 = _mm_shuffle_epi32(VFilter, _MM_SHUFFLE(3, 3, 3, 3));
+      __m128i col0, col1, col2, col3;
+      {
+        //load pixels
+        __m128i src  = transpose3_0;
+        // extract the ones used for first column
+        __m128i src0123 = _mm_shuffle_epi8(src, mask0123);
+        __m128i src4567 = _mm_shuffle_epi8(src, mask4567);
+        __m128i src01_16 = _mm_unpacklo_epi8(src0123, zero);
+        __m128i src23_16 = _mm_unpackhi_epi8(src0123, zero);
+        __m128i src45_16 = _mm_unpacklo_epi8(src4567, zero);
+        __m128i src67_16 = _mm_unpackhi_epi8(src4567, zero);
+        // multiply accumulate them
+        __m128i mad01 = _mm_madd_epi16(src01_16, fil01);
+        __m128i mad23 = _mm_madd_epi16(src23_16, fil23);
+        __m128i mad45 = _mm_madd_epi16(src45_16, fil45);
+        __m128i mad67 = _mm_madd_epi16(src67_16, fil67);
+        __m128i mad0123 = _mm_add_epi32(mad01, mad23);
+        __m128i mad4567 = _mm_add_epi32(mad45, mad67);
+        __m128i mad_all = _mm_add_epi32(mad0123, mad4567);
+        mad_all = _mm_add_epi32(mad_all, rounding);
+        mad_all = _mm_srai_epi32(mad_all, VP8_FILTER_SHIFT);
+        mad_all = _mm_packs_epi32(mad_all, mad_all);
+        col0 = _mm_packus_epi16(mad_all, mad_all);
+      }
+      {
+        //load pixels
+        __m128i src  = transpose3_1;
+        // extract the ones used for first column
+        __m128i src0123 = _mm_shuffle_epi8(src, mask0123);
+        __m128i src4567 = _mm_shuffle_epi8(src, mask4567);
+        __m128i src01_16 = _mm_unpacklo_epi8(src0123, zero);
+        __m128i src23_16 = _mm_unpackhi_epi8(src0123, zero);
+        __m128i src45_16 = _mm_unpacklo_epi8(src4567, zero);
+        __m128i src67_16 = _mm_unpackhi_epi8(src4567, zero);
+        // multiply accumulate them
+        __m128i mad01 = _mm_madd_epi16(src01_16, fil01);
+        __m128i mad23 = _mm_madd_epi16(src23_16, fil23);
+        __m128i mad45 = _mm_madd_epi16(src45_16, fil45);
+        __m128i mad67 = _mm_madd_epi16(src67_16, fil67);
+        __m128i mad0123 = _mm_add_epi32(mad01, mad23);
+        __m128i mad4567 = _mm_add_epi32(mad45, mad67);
+        __m128i mad_all = _mm_add_epi32(mad0123, mad4567);
+        mad_all = _mm_add_epi32(mad_all, rounding);
+        mad_all = _mm_srai_epi32(mad_all, VP8_FILTER_SHIFT);
+        mad_all = _mm_packs_epi32(mad_all, mad_all);
+        col1 = _mm_packus_epi16(mad_all, mad_all);
+      }
+      {
+        //load pixels
+        __m128i src  = transpose3_2;
+        // extract the ones used for first column
+        __m128i src0123 = _mm_shuffle_epi8(src, mask0123);
+        __m128i src4567 = _mm_shuffle_epi8(src, mask4567);
+        __m128i src01_16 = _mm_unpacklo_epi8(src0123, zero);
+        __m128i src23_16 = _mm_unpackhi_epi8(src0123, zero);
+        __m128i src45_16 = _mm_unpacklo_epi8(src4567, zero);
+        __m128i src67_16 = _mm_unpackhi_epi8(src4567, zero);
+        // multiply accumulate them
+        __m128i mad01 = _mm_madd_epi16(src01_16, fil01);
+        __m128i mad23 = _mm_madd_epi16(src23_16, fil23);
+        __m128i mad45 = _mm_madd_epi16(src45_16, fil45);
+        __m128i mad67 = _mm_madd_epi16(src67_16, fil67);
+        __m128i mad0123 = _mm_add_epi32(mad01, mad23);
+        __m128i mad4567 = _mm_add_epi32(mad45, mad67);
+        __m128i mad_all = _mm_add_epi32(mad0123, mad4567);
+        mad_all = _mm_add_epi32(mad_all, rounding);
+        mad_all = _mm_srai_epi32(mad_all, VP8_FILTER_SHIFT);
+        mad_all = _mm_packs_epi32(mad_all, mad_all);
+        col2 = _mm_packus_epi16(mad_all, mad_all);
+      }
+      {
+        //load pixels
+        __m128i src  = transpose3_3;
+        // extract the ones used for first column
+        __m128i src0123 = _mm_shuffle_epi8(src, mask0123);
+        __m128i src4567 = _mm_shuffle_epi8(src, mask4567);
+        __m128i src01_16 = _mm_unpacklo_epi8(src0123, zero);
+        __m128i src23_16 = _mm_unpackhi_epi8(src0123, zero);
+        __m128i src45_16 = _mm_unpacklo_epi8(src4567, zero);
+        __m128i src67_16 = _mm_unpackhi_epi8(src4567, zero);
+        // multiply accumulate them
+        __m128i mad01 = _mm_madd_epi16(src01_16, fil01);
+        __m128i mad23 = _mm_madd_epi16(src23_16, fil23);
+        __m128i mad45 = _mm_madd_epi16(src45_16, fil45);
+        __m128i mad67 = _mm_madd_epi16(src67_16, fil67);
+        __m128i mad0123 = _mm_add_epi32(mad01, mad23);
+        __m128i mad4567 = _mm_add_epi32(mad45, mad67);
+        __m128i mad_all = _mm_add_epi32(mad0123, mad4567);
+        mad_all = _mm_add_epi32(mad_all, rounding);
+        mad_all = _mm_srai_epi32(mad_all, VP8_FILTER_SHIFT);
+        mad_all = _mm_packs_epi32(mad_all, mad_all);
+        col3 = _mm_packus_epi16(mad_all, mad_all);
+      }
+      {
+        __m128i col01 = _mm_unpacklo_epi8(col0, col1);
+        __m128i col23 = _mm_unpacklo_epi8(col2, col3);
+        __m128i col0123 = _mm_unpacklo_epi16(col01, col23);
+        //TODO(cd): look into Ronald's comment:
+        //    Future suggestion: I believe here, too, you can merge the
+        //    packs_epi32() and pacus_epi16() for the 4 cols above, so that
+        //    you get the data in a single register, and then use pshufb
+        //    (shuffle_epi8()) instead of the unpacks here. Should be
+        //    2+3+2 instructions faster.
+        *((unsigned int *)&dst_ptr[dst_stride * 0]) =
+            _mm_extract_epi32(col0123, 0);
+        *((unsigned int *)&dst_ptr[dst_stride * 1]) =
+            _mm_extract_epi32(col0123, 1);
+        *((unsigned int *)&dst_ptr[dst_stride * 2]) =
+            _mm_extract_epi32(col0123, 2);
+        *((unsigned int *)&dst_ptr[dst_stride * 3]) =
+            _mm_extract_epi32(col0123, 3);
+      }
+    }
+  }
+}
+
+void vp8_filter_block2d_8x4_8_sse4_1
+(
+ const unsigned char *src_ptr, const unsigned int src_stride,
+ const short *HFilter_aligned16, const short *VFilter_aligned16,
+ unsigned char *dst_ptr, unsigned int dst_stride
+) {
+  int j;
+  for (j=0; j<8; j+=4) {
+    vp8_filter_block2d_4x4_8_sse4_1(src_ptr + j, src_stride,
+                                    HFilter_aligned16, VFilter_aligned16,
+                                    dst_ptr + j, dst_stride);
+  }
+}
+
+void vp8_filter_block2d_8x8_8_sse4_1
+(
+ const unsigned char *src_ptr, const unsigned int src_stride,
+ const short *HFilter_aligned16, const short *VFilter_aligned16,
+ unsigned char *dst_ptr, unsigned int dst_stride
+) {
+  int i, j;
+  for (i=0; i<8; i+=4) {
+    for (j=0; j<8; j+=4) {
+      vp8_filter_block2d_4x4_8_sse4_1(src_ptr + j + i*src_stride, src_stride,
+                                      HFilter_aligned16, VFilter_aligned16,
+                                      dst_ptr + j + i*dst_stride, dst_stride);
+    }
+  }
+}
+
+void vp8_filter_block2d_16x16_8_sse4_1
+(
+ const unsigned char *src_ptr, const unsigned int src_stride,
+ const short *HFilter_aligned16, const short *VFilter_aligned16,
+ unsigned char *dst_ptr, unsigned int dst_stride
+) {
+  int i, j;
+  for (i=0; i<16; i+=4) {
+    for (j=0; j<16; j+=4) {
+      vp8_filter_block2d_4x4_8_sse4_1(src_ptr + j + i*src_stride, src_stride,
+                                      HFilter_aligned16, VFilter_aligned16,
+                                      dst_ptr + j + i*dst_stride, dst_stride);
+    }
+  }
+}
diff --git a/vp8/vp8_common.mk b/vp8/vp8_common.mk
index 8c9addc78..3595a851a 100644
--- a/vp8/vp8_common.mk
+++ b/vp8/vp8_common.mk
@@ -114,6 +114,11 @@ VP8_COMMON_SRCS-yes += common/maskingmv.c
 VP8_COMMON_SRCS-$(HAVE_SSE3) += common/x86/mask_sse3.asm
 endif
 
+VP8_COMMON_SRCS-$(HAVE_SSE4_1) += common/x86/filter_sse4.c
+ifeq ($(HAVE_SSE4_1),yes)
+vp8/common/x86/filter_sse4.c.o: CFLAGS += -msse4
+endif
+
 VP8_COMMON_SRCS-$(ARCH_ARM)  += common/arm/arm_systemdependent.c
 VP8_COMMON_SRCS-$(ARCH_ARM)  += common/arm/bilinearfilter_arm.c
 VP8_COMMON_SRCS-$(ARCH_ARM)  += common/arm/bilinearfilter_arm.h