Add rectangular block size variance/sad functions.

With this, the RD loop properly supports rectangular blocks.

Change-Id: Iece79048fb4e84741ee1ada982da129a7bf00470

6 files changed, 382 insertions, 44 deletions

diff --git a/vp9/common/vp9_rtcd_defs.sh b/vp9/common/vp9_rtcd_defs.sh
index 8ce9574c2..9a20aa6ad 100644
--- a/vp9/common/vp9_rtcd_defs.sh
+++ b/vp9/common/vp9_rtcd_defs.sh
@@ -376,6 +376,20 @@ if [ "$CONFIG_VP9_ENCODER" = "yes" ]; then
 # variance
 [ $arch = "x86_64" ] && mmx_x86_64=mmx && sse2_x86_64=sse2
 
+#if CONFIG_SBSEGMENT
+prototype unsigned int vp9_variance32x16 "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse"
+specialize vp9_variance32x16
+
+prototype unsigned int vp9_variance16x32 "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse"
+specialize vp9_variance16x32
+
+prototype unsigned int vp9_variance64x32 "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse"
+specialize vp9_variance64x32
+
+prototype unsigned int vp9_variance32x64 "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse"
+specialize vp9_variance32x64
+#endif
+
 prototype unsigned int vp9_variance32x32 "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse"
 specialize vp9_variance32x32
 
@@ -410,6 +424,20 @@ vp9_variance4x4_mmx=vp9_variance4x4_mmx
 prototype unsigned int vp9_sub_pixel_variance64x64 "const uint8_t *src_ptr, int source_stride, int xoffset, int  yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse"
 specialize vp9_sub_pixel_variance64x64 sse2
 
+#if CONFIG_SBSEGMENT
+prototype unsigned int vp9_sub_pixel_variance32x64 "const uint8_t *src_ptr, int source_stride, int xoffset, int  yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse"
+specialize vp9_sub_pixel_variance32x64
+
+prototype unsigned int vp9_sub_pixel_variance64x32 "const uint8_t *src_ptr, int source_stride, int xoffset, int  yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse"
+specialize vp9_sub_pixel_variance64x32
+
+prototype unsigned int vp9_sub_pixel_variance32x16 "const uint8_t *src_ptr, int source_stride, int xoffset, int  yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse"
+specialize vp9_sub_pixel_variance32x16
+
+prototype unsigned int vp9_sub_pixel_variance16x32 "const uint8_t *src_ptr, int source_stride, int xoffset, int  yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse"
+specialize vp9_sub_pixel_variance16x32
+#endif
+
 prototype unsigned int vp9_sub_pixel_variance32x32 "const uint8_t *src_ptr, int source_stride, int xoffset, int  yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse"
 specialize vp9_sub_pixel_variance32x32 sse2
 
@@ -436,6 +464,20 @@ vp9_sub_pixel_variance4x4_sse2=vp9_sub_pixel_variance4x4_wmt
 prototype unsigned int vp9_sad64x64 "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int  ref_stride, unsigned int max_sad"
 specialize vp9_sad64x64 sse2
 
+#if CONFIG_SBSEGMENT
+prototype unsigned int vp9_sad32x64 "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int max_sad"
+specialize vp9_sad32x64
+
+prototype unsigned int vp9_sad64x32 "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int max_sad"
+specialize vp9_sad64x32
+
+prototype unsigned int vp9_sad32x16 "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int max_sad"
+specialize vp9_sad32x16
+
+prototype unsigned int vp9_sad16x32 "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int max_sad"
+specialize vp9_sad16x32
+#endif
+
 prototype unsigned int vp9_sad32x32 "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int  ref_stride, unsigned int max_sad"
 specialize vp9_sad32x32 sse2
 
@@ -529,6 +571,20 @@ specialize vp9_sad4x4x8 sse4
 prototype void vp9_sad64x64x4d "const uint8_t *src_ptr, int  src_stride, const uint8_t* const ref_ptr[], int  ref_stride, unsigned int *sad_array"
 specialize vp9_sad64x64x4d sse2
 
+#if CONFIG_SBSEGMENT
+prototype void vp9_sad32x64x4d "const uint8_t *src_ptr, int  src_stride, const uint8_t* const ref_ptr[], int  ref_stride, unsigned int *sad_array"
+specialize vp9_sad32x64x4d
+
+prototype void vp9_sad64x32x4d "const uint8_t *src_ptr, int  src_stride, const uint8_t* const ref_ptr[], int  ref_stride, unsigned int *sad_array"
+specialize vp9_sad64x32x4d
+
+prototype void vp9_sad32x16x4d "const uint8_t *src_ptr, int  src_stride, const uint8_t* const ref_ptr[], int  ref_stride, unsigned int *sad_array"
+specialize vp9_sad32x16x4d
+
+prototype void vp9_sad16x32x4d "const uint8_t *src_ptr, int  src_stride, const uint8_t* const ref_ptr[], int  ref_stride, unsigned int *sad_array"
+specialize vp9_sad16x32x4d
+#endif
+
 prototype void vp9_sad32x32x4d "const uint8_t *src_ptr, int  src_stride, const uint8_t* const ref_ptr[], int  ref_stride, unsigned int *sad_array"
 specialize vp9_sad32x32x4d sse2
 
diff --git a/vp9/encoder/vp9_onyx_if.c b/vp9/encoder/vp9_onyx_if.c
index 05977b3f5..b120b722e 100644
--- a/vp9/encoder/vp9_onyx_if.c
+++ b/vp9/encoder/vp9_onyx_if.c
@@ -1560,6 +1560,27 @@ VP9_PTR vp9_create_compressor(VP9_CONFIG *oxcf) {
     cpi->fn_ptr[BT].sdx8f          = SDX8F; \
     cpi->fn_ptr[BT].sdx4df         = SDX4DF;
 
+#if CONFIG_SBSEGMENT
+  BFP(BLOCK_32X16, vp9_sad32x16, vp9_variance32x16, vp9_sub_pixel_variance32x16,
+      NULL, NULL,
+      NULL, NULL, NULL,
+      vp9_sad32x16x4d)
+
+  BFP(BLOCK_16X32, vp9_sad16x32, vp9_variance16x32, vp9_sub_pixel_variance16x32,
+      NULL, NULL,
+      NULL, NULL, NULL,
+      vp9_sad16x32x4d)
+
+  BFP(BLOCK_64X32, vp9_sad64x32, vp9_variance64x32, vp9_sub_pixel_variance64x32,
+      NULL, NULL,
+      NULL, NULL, NULL,
+      vp9_sad64x32x4d)
+
+  BFP(BLOCK_32X64, vp9_sad32x64, vp9_variance32x64, vp9_sub_pixel_variance32x64,
+      NULL, NULL,
+      NULL, NULL, NULL,
+      vp9_sad32x64x4d)
+#endif
 
   BFP(BLOCK_32X32, vp9_sad32x32, vp9_variance32x32, vp9_sub_pixel_variance32x32,
       vp9_variance_halfpixvar32x32_h, vp9_variance_halfpixvar32x32_v,
diff --git a/vp9/encoder/vp9_onyx_int.h b/vp9/encoder/vp9_onyx_int.h
index 130bc85f3..2ffdfc29a 100644
--- a/vp9/encoder/vp9_onyx_int.h
+++ b/vp9/encoder/vp9_onyx_int.h
@@ -281,6 +281,12 @@ enum BlockSize {
   BLOCK_16X16,
   BLOCK_MAX_SEGMENTS,
   BLOCK_32X32 = BLOCK_MAX_SEGMENTS,
+#if CONFIG_SBSEGMENT
+  BLOCK_32X16,
+  BLOCK_16X32,
+  BLOCK_64X32,
+  BLOCK_32X64,
+#endif
   BLOCK_64X64,
   BLOCK_MAX_SB_SEGMENTS,
 };
diff --git a/vp9/encoder/vp9_rdopt.c b/vp9/encoder/vp9_rdopt.c
index 4d1cf9733..9d1cada00 100644
--- a/vp9/encoder/vp9_rdopt.c
+++ b/vp9/encoder/vp9_rdopt.c
@@ -2813,7 +2813,15 @@ static void model_rd_from_var_lapndz(int var, int n, int qstep,
 static enum BlockSize y_to_uv_block_size(enum BlockSize bs) {
   switch (bs) {
     case BLOCK_64X64: return BLOCK_32X32;
+#if CONFIG_SBSEGMENT
+    case BLOCK_64X32: return BLOCK_32X16;
+    case BLOCK_32X64: return BLOCK_16X32;
+#endif
     case BLOCK_32X32: return BLOCK_16X16;
+#if CONFIG_SBSEGMENT
+    case BLOCK_32X16: return BLOCK_16X8;
+    case BLOCK_16X32: return BLOCK_8X16;
+#endif
     case BLOCK_16X16: return BLOCK_8X8;
     default:
       assert(0);
@@ -2824,7 +2832,15 @@ static enum BlockSize y_to_uv_block_size(enum BlockSize bs) {
 static enum BlockSize y_bsizet_to_block_size(BLOCK_SIZE_TYPE bs) {
   switch (bs) {
     case BLOCK_SIZE_SB64X64: return BLOCK_64X64;
+#if CONFIG_SBSEGMENT
+    case BLOCK_SIZE_SB64X32: return BLOCK_64X32;
+    case BLOCK_SIZE_SB32X64: return BLOCK_32X64;
+#endif
     case BLOCK_SIZE_SB32X32: return BLOCK_32X32;
+#if CONFIG_SBSEGMENT
+    case BLOCK_SIZE_SB32X16: return BLOCK_32X16;
+    case BLOCK_SIZE_SB16X32: return BLOCK_16X32;
+#endif
     case BLOCK_SIZE_MB16X16: return BLOCK_16X16;
     default:
       assert(0);
diff --git a/vp9/encoder/vp9_sad_c.c b/vp9/encoder/vp9_sad_c.c
index af5526dce..dbadaea74 100644
--- a/vp9/encoder/vp9_sad_c.c
+++ b/vp9/encoder/vp9_sad_c.c
@@ -23,6 +23,54 @@ unsigned int vp9_sad64x64_c(const uint8_t *src_ptr,
   return sad_mx_n_c(src_ptr, src_stride, ref_ptr, ref_stride, 64, 64);
 }
 
+#if CONFIG_SBSEGMENT
+unsigned int vp9_sad64x32_c(const uint8_t *src_ptr,
+                            int  src_stride,
+                            const uint8_t *ref_ptr,
+                            int  ref_stride,
+                            unsigned int max_sad) {
+  return sad_mx_n_c(src_ptr, src_stride, ref_ptr, ref_stride, 64, 32);
+}
+
+void vp9_sad64x32x4d_c(const uint8_t *src_ptr,
+                       int  src_stride,
+                       const uint8_t* const ref_ptr[],
+                       int  ref_stride,
+                       unsigned int *sad_array) {
+  sad_array[0] = vp9_sad64x32(src_ptr, src_stride,
+                              ref_ptr[0], ref_stride, 0x7fffffff);
+  sad_array[1] = vp9_sad64x32(src_ptr, src_stride,
+                              ref_ptr[1], ref_stride, 0x7fffffff);
+  sad_array[2] = vp9_sad64x32(src_ptr, src_stride,
+                              ref_ptr[2], ref_stride, 0x7fffffff);
+  sad_array[3] = vp9_sad64x32(src_ptr, src_stride,
+                              ref_ptr[3], ref_stride, 0x7fffffff);
+}
+
+unsigned int vp9_sad32x64_c(const uint8_t *src_ptr,
+                            int  src_stride,
+                            const uint8_t *ref_ptr,
+                            int  ref_stride,
+                            unsigned int max_sad) {
+  return sad_mx_n_c(src_ptr, src_stride, ref_ptr, ref_stride, 32, 64);
+}
+
+void vp9_sad32x64x4d_c(const uint8_t *src_ptr,
+                       int  src_stride,
+                       const uint8_t* const ref_ptr[],
+                       int  ref_stride,
+                       unsigned int *sad_array) {
+  sad_array[0] = vp9_sad32x64(src_ptr, src_stride,
+                              ref_ptr[0], ref_stride, 0x7fffffff);
+  sad_array[1] = vp9_sad32x64(src_ptr, src_stride,
+                              ref_ptr[1], ref_stride, 0x7fffffff);
+  sad_array[2] = vp9_sad32x64(src_ptr, src_stride,
+                              ref_ptr[2], ref_stride, 0x7fffffff);
+  sad_array[3] = vp9_sad32x64(src_ptr, src_stride,
+                              ref_ptr[3], ref_stride, 0x7fffffff);
+}
+#endif
+
 unsigned int vp9_sad32x32_c(const uint8_t *src_ptr,
                             int  src_stride,
                             const uint8_t *ref_ptr,
@@ -31,6 +79,54 @@ unsigned int vp9_sad32x32_c(const uint8_t *src_ptr,
   return sad_mx_n_c(src_ptr, src_stride, ref_ptr, ref_stride, 32, 32);
 }
 
+#if CONFIG_SBSEGMENT
+unsigned int vp9_sad32x16_c(const uint8_t *src_ptr,
+                            int   src_stride,
+                            const uint8_t *ref_ptr,
+                            int   ref_stride,
+                            unsigned int max_sad) {
+  return sad_mx_n_c(src_ptr, src_stride, ref_ptr, ref_stride, 32, 16);
+}
+
+void vp9_sad32x16x4d_c(const uint8_t *src_ptr,
+                       int  src_stride,
+                       const uint8_t* const ref_ptr[],
+                       int  ref_stride,
+                       unsigned int *sad_array) {
+  sad_array[0] = vp9_sad32x16(src_ptr, src_stride,
+                              ref_ptr[0], ref_stride, 0x7fffffff);
+  sad_array[1] = vp9_sad32x16(src_ptr, src_stride,
+                              ref_ptr[1], ref_stride, 0x7fffffff);
+  sad_array[2] = vp9_sad32x16(src_ptr, src_stride,
+                              ref_ptr[2], ref_stride, 0x7fffffff);
+  sad_array[3] = vp9_sad32x16(src_ptr, src_stride,
+                              ref_ptr[3], ref_stride, 0x7fffffff);
+}
+
+unsigned int vp9_sad16x32_c(const uint8_t *src_ptr,
+                            int   src_stride,
+                            const uint8_t *ref_ptr,
+                            int   ref_stride,
+                            unsigned int max_sad) {
+  return sad_mx_n_c(src_ptr, src_stride, ref_ptr, ref_stride, 16, 32);
+}
+
+void vp9_sad16x32x4d_c(const uint8_t *src_ptr,
+                       int  src_stride,
+                       const uint8_t* const ref_ptr[],
+                       int  ref_stride,
+                       unsigned int *sad_array) {
+  sad_array[0] = vp9_sad16x32(src_ptr, src_stride,
+                              ref_ptr[0], ref_stride, 0x7fffffff);
+  sad_array[1] = vp9_sad16x32(src_ptr, src_stride,
+                              ref_ptr[1], ref_stride, 0x7fffffff);
+  sad_array[2] = vp9_sad16x32(src_ptr, src_stride,
+                              ref_ptr[2], ref_stride, 0x7fffffff);
+  sad_array[3] = vp9_sad16x32(src_ptr, src_stride,
+                              ref_ptr[3], ref_stride, 0x7fffffff);
+}
+#endif
+
 unsigned int vp9_sad16x16_c(const uint8_t *src_ptr,
                             int  src_stride,
                             const uint8_t *ref_ptr,
diff --git a/vp9/encoder/vp9_variance_c.c b/vp9/encoder/vp9_variance_c.c
index d07a65b45..fb66f4f0b 100644
--- a/vp9/encoder/vp9_variance_c.c
+++ b/vp9/encoder/vp9_variance_c.c
@@ -24,6 +24,144 @@ unsigned int vp9_get_mb_ss_c(const int16_t *src_ptr) {
   return sum;
 }
 
+#if CONFIG_SBSEGMENT
+unsigned int vp9_variance64x32_c(const uint8_t *src_ptr,
+                                 int  source_stride,
+                                 const uint8_t *ref_ptr,
+                                 int  recon_stride,
+                                 unsigned int *sse) {
+  unsigned int var;
+  int avg;
+
+  variance(src_ptr, source_stride, ref_ptr, recon_stride, 64, 32, &var, &avg);
+  *sse = var;
+  return (var - (((int64_t)avg * avg) >> 11));
+}
+
+unsigned int vp9_sub_pixel_variance64x32_c(const uint8_t *src_ptr,
+                                           int  src_pixels_per_line,
+                                           int  xoffset,
+                                           int  yoffset,
+                                           const uint8_t *dst_ptr,
+                                           int dst_pixels_per_line,
+                                           unsigned int *sse) {
+  uint16_t fdata3[65 * 64];  // Temp data bufffer used in filtering
+  uint8_t temp2[68 * 64];
+  const int16_t *hfilter, *vfilter;
+
+  hfilter = VP9_BILINEAR_FILTERS_2TAP(xoffset);
+  vfilter = VP9_BILINEAR_FILTERS_2TAP(yoffset);
+
+  var_filter_block2d_bil_first_pass(src_ptr, fdata3, src_pixels_per_line,
+                                    1, 33, 64, hfilter);
+  var_filter_block2d_bil_second_pass(fdata3, temp2, 64, 64, 32, 64, vfilter);
+
+  return vp9_variance64x32_c(temp2, 64, dst_ptr, dst_pixels_per_line, sse);
+}
+
+unsigned int vp9_variance32x64_c(const uint8_t *src_ptr,
+                                 int  source_stride,
+                                 const uint8_t *ref_ptr,
+                                 int  recon_stride,
+                                 unsigned int *sse) {
+  unsigned int var;
+  int avg;
+
+  variance(src_ptr, source_stride, ref_ptr, recon_stride, 32, 64, &var, &avg);
+  *sse = var;
+  return (var - (((int64_t)avg * avg) >> 11));
+}
+
+unsigned int vp9_sub_pixel_variance32x64_c(const uint8_t *src_ptr,
+                                           int  src_pixels_per_line,
+                                           int  xoffset,
+                                           int  yoffset,
+                                           const uint8_t *dst_ptr,
+                                           int dst_pixels_per_line,
+                                           unsigned int *sse) {
+  uint16_t fdata3[65 * 64];  // Temp data bufffer used in filtering
+  uint8_t temp2[68 * 64];
+  const int16_t *hfilter, *vfilter;
+
+  hfilter = VP9_BILINEAR_FILTERS_2TAP(xoffset);
+  vfilter = VP9_BILINEAR_FILTERS_2TAP(yoffset);
+
+  var_filter_block2d_bil_first_pass(src_ptr, fdata3, src_pixels_per_line,
+                                    1, 65, 32, hfilter);
+  var_filter_block2d_bil_second_pass(fdata3, temp2, 64, 64, 64, 32, vfilter);
+
+  return vp9_variance32x64_c(temp2, 64, dst_ptr, dst_pixels_per_line, sse);
+}
+
+unsigned int vp9_variance32x16_c(const uint8_t *src_ptr,
+                                 int  source_stride,
+                                 const uint8_t *ref_ptr,
+                                 int  recon_stride,
+                                 unsigned int *sse) {
+  unsigned int var;
+  int avg;
+
+  variance(src_ptr, source_stride, ref_ptr, recon_stride, 32, 16, &var, &avg);
+  *sse = var;
+  return (var - (((int64_t)avg * avg) >> 9));
+}
+
+unsigned int vp9_sub_pixel_variance32x16_c(const uint8_t *src_ptr,
+                                           int  src_pixels_per_line,
+                                           int  xoffset,
+                                           int  yoffset,
+                                           const uint8_t *dst_ptr,
+                                           int dst_pixels_per_line,
+                                           unsigned int *sse) {
+  uint16_t fdata3[33 * 32];  // Temp data bufffer used in filtering
+  uint8_t temp2[36 * 32];
+  const int16_t *hfilter, *vfilter;
+
+  hfilter = VP9_BILINEAR_FILTERS_2TAP(xoffset);
+  vfilter = VP9_BILINEAR_FILTERS_2TAP(yoffset);
+
+  var_filter_block2d_bil_first_pass(src_ptr, fdata3, src_pixels_per_line,
+                                    1, 17, 32, hfilter);
+  var_filter_block2d_bil_second_pass(fdata3, temp2, 32, 32, 16, 32, vfilter);
+
+  return vp9_variance32x16_c(temp2, 32, dst_ptr, dst_pixels_per_line, sse);
+}
+
+unsigned int vp9_variance16x32_c(const uint8_t *src_ptr,
+                                 int  source_stride,
+                                 const uint8_t *ref_ptr,
+                                 int  recon_stride,
+                                 unsigned int *sse) {
+  unsigned int var;
+  int avg;
+
+  variance(src_ptr, source_stride, ref_ptr, recon_stride, 16, 32, &var, &avg);
+  *sse = var;
+  return (var - (((int64_t)avg * avg) >> 9));
+}
+
+unsigned int vp9_sub_pixel_variance16x32_c(const uint8_t *src_ptr,
+                                           int  src_pixels_per_line,
+                                           int  xoffset,
+                                           int  yoffset,
+                                           const uint8_t *dst_ptr,
+                                           int dst_pixels_per_line,
+                                           unsigned int *sse) {
+  uint16_t fdata3[33 * 32];  // Temp data bufffer used in filtering
+  uint8_t temp2[36 * 32];
+  const int16_t *hfilter, *vfilter;
+
+  hfilter = VP9_BILINEAR_FILTERS_2TAP(xoffset);
+  vfilter = VP9_BILINEAR_FILTERS_2TAP(yoffset);
+
+  var_filter_block2d_bil_first_pass(src_ptr, fdata3, src_pixels_per_line,
+                                    1, 33, 16, hfilter);
+  var_filter_block2d_bil_second_pass(fdata3, temp2, 32, 32, 32, 16, vfilter);
+
+  return vp9_variance16x32_c(temp2, 32, dst_ptr, dst_pixels_per_line, sse);
+}
+#endif
+
 unsigned int vp9_variance64x64_c(const uint8_t *src_ptr,
                                  int  source_stride,
                                  const uint8_t *ref_ptr,
@@ -139,17 +277,18 @@ unsigned int vp9_sub_pixel_variance4x4_c(const uint8_t *src_ptr,
                                          int dst_pixels_per_line,
                                          unsigned int *sse) {
   uint8_t temp2[20 * 16];
-  const int16_t *HFilter, *VFilter;
-  uint16_t FData3[5 * 4];  // Temp data bufffer used in filtering
+  const int16_t *hfilter, *vfilter;
+  uint16_t fdata3[5 * 4];  // Temp data bufffer used in filtering
 
-  HFilter = VP9_BILINEAR_FILTERS_2TAP(xoffset);
-  VFilter = VP9_BILINEAR_FILTERS_2TAP(yoffset);
+  hfilter = VP9_BILINEAR_FILTERS_2TAP(xoffset);
+  vfilter = VP9_BILINEAR_FILTERS_2TAP(yoffset);
 
   // First filter 1d Horizontal
-  var_filter_block2d_bil_first_pass(src_ptr, FData3, src_pixels_per_line, 1, 5, 4, HFilter);
+  var_filter_block2d_bil_first_pass(src_ptr, fdata3, src_pixels_per_line,
+                                    1, 5, 4, hfilter);
 
   // Now filter Verticaly
-  var_filter_block2d_bil_second_pass(FData3, temp2, 4,  4,  4,  4, VFilter);
+  var_filter_block2d_bil_second_pass(fdata3, temp2, 4,  4,  4,  4, vfilter);
 
   return vp9_variance4x4_c(temp2, 4, dst_ptr, dst_pixels_per_line, sse);
 }
@@ -162,15 +301,16 @@ unsigned int vp9_sub_pixel_variance8x8_c(const uint8_t *src_ptr,
                                          const uint8_t *dst_ptr,
                                          int dst_pixels_per_line,
                                          unsigned int *sse) {
-  uint16_t FData3[9 * 8];  // Temp data bufffer used in filtering
+  uint16_t fdata3[9 * 8];  // Temp data bufffer used in filtering
   uint8_t temp2[20 * 16];
-  const int16_t *HFilter, *VFilter;
+  const int16_t *hfilter, *vfilter;
 
-  HFilter = VP9_BILINEAR_FILTERS_2TAP(xoffset);
-  VFilter = VP9_BILINEAR_FILTERS_2TAP(yoffset);
+  hfilter = VP9_BILINEAR_FILTERS_2TAP(xoffset);
+  vfilter = VP9_BILINEAR_FILTERS_2TAP(yoffset);
 
-  var_filter_block2d_bil_first_pass(src_ptr, FData3, src_pixels_per_line, 1, 9, 8, HFilter);
-  var_filter_block2d_bil_second_pass(FData3, temp2, 8, 8, 8, 8, VFilter);
+  var_filter_block2d_bil_first_pass(src_ptr, fdata3, src_pixels_per_line,
+                                    1, 9, 8, hfilter);
+  var_filter_block2d_bil_second_pass(fdata3, temp2, 8, 8, 8, 8, vfilter);
 
   return vp9_variance8x8_c(temp2, 8, dst_ptr, dst_pixels_per_line, sse);
 }
@@ -182,15 +322,16 @@ unsigned int vp9_sub_pixel_variance16x16_c(const uint8_t *src_ptr,
                                            const uint8_t *dst_ptr,
                                            int dst_pixels_per_line,
                                            unsigned int *sse) {
-  uint16_t FData3[17 * 16];  // Temp data bufffer used in filtering
+  uint16_t fdata3[17 * 16];  // Temp data bufffer used in filtering
   uint8_t temp2[20 * 16];
-  const int16_t *HFilter, *VFilter;
+  const int16_t *hfilter, *vfilter;
 
-  HFilter = VP9_BILINEAR_FILTERS_2TAP(xoffset);
-  VFilter = VP9_BILINEAR_FILTERS_2TAP(yoffset);
+  hfilter = VP9_BILINEAR_FILTERS_2TAP(xoffset);
+  vfilter = VP9_BILINEAR_FILTERS_2TAP(yoffset);
 
-  var_filter_block2d_bil_first_pass(src_ptr, FData3, src_pixels_per_line, 1, 17, 16, HFilter);
-  var_filter_block2d_bil_second_pass(FData3, temp2, 16, 16, 16, 16, VFilter);
+  var_filter_block2d_bil_first_pass(src_ptr, fdata3, src_pixels_per_line,
+                                    1, 17, 16, hfilter);
+  var_filter_block2d_bil_second_pass(fdata3, temp2, 16, 16, 16, 16, vfilter);
 
   return vp9_variance16x16_c(temp2, 16, dst_ptr, dst_pixels_per_line, sse);
 }
@@ -202,16 +343,16 @@ unsigned int vp9_sub_pixel_variance64x64_c(const uint8_t *src_ptr,
                                            const uint8_t *dst_ptr,
                                            int dst_pixels_per_line,
                                            unsigned int *sse) {
-  uint16_t FData3[65 * 64];  // Temp data bufffer used in filtering
+  uint16_t fdata3[65 * 64];  // Temp data bufffer used in filtering
   uint8_t temp2[68 * 64];
-  const int16_t *HFilter, *VFilter;
+  const int16_t *hfilter, *vfilter;
 
-  HFilter = VP9_BILINEAR_FILTERS_2TAP(xoffset);
-  VFilter = VP9_BILINEAR_FILTERS_2TAP(yoffset);
+  hfilter = VP9_BILINEAR_FILTERS_2TAP(xoffset);
+  vfilter = VP9_BILINEAR_FILTERS_2TAP(yoffset);
 
-  var_filter_block2d_bil_first_pass(src_ptr, FData3, src_pixels_per_line,
-                                    1, 65, 64, HFilter);
-  var_filter_block2d_bil_second_pass(FData3, temp2, 64, 64, 64, 64, VFilter);
+  var_filter_block2d_bil_first_pass(src_ptr, fdata3, src_pixels_per_line,
+                                    1, 65, 64, hfilter);
+  var_filter_block2d_bil_second_pass(fdata3, temp2, 64, 64, 64, 64, vfilter);
 
   return vp9_variance64x64_c(temp2, 64, dst_ptr, dst_pixels_per_line, sse);
 }
@@ -223,15 +364,16 @@ unsigned int vp9_sub_pixel_variance32x32_c(const uint8_t *src_ptr,
                                            const uint8_t *dst_ptr,
                                            int dst_pixels_per_line,
                                            unsigned int *sse) {
-  uint16_t FData3[33 * 32];  // Temp data bufffer used in filtering
+  uint16_t fdata3[33 * 32];  // Temp data bufffer used in filtering
   uint8_t temp2[36 * 32];
-  const int16_t *HFilter, *VFilter;
+  const int16_t *hfilter, *vfilter;
 
-  HFilter = VP9_BILINEAR_FILTERS_2TAP(xoffset);
-  VFilter = VP9_BILINEAR_FILTERS_2TAP(yoffset);
+  hfilter = VP9_BILINEAR_FILTERS_2TAP(xoffset);
+  vfilter = VP9_BILINEAR_FILTERS_2TAP(yoffset);
 
-  var_filter_block2d_bil_first_pass(src_ptr, FData3, src_pixels_per_line, 1, 33, 32, HFilter);
-  var_filter_block2d_bil_second_pass(FData3, temp2, 32, 32, 32, 32, VFilter);
+  var_filter_block2d_bil_first_pass(src_ptr, fdata3, src_pixels_per_line,
+                                    1, 33, 32, hfilter);
+  var_filter_block2d_bil_second_pass(fdata3, temp2, 32, 32, 32, 32, vfilter);
 
   return vp9_variance32x32_c(temp2, 32, dst_ptr, dst_pixels_per_line, sse);
 }
@@ -363,15 +505,16 @@ unsigned int vp9_sub_pixel_variance16x8_c(const uint8_t *src_ptr,
                                           const uint8_t *dst_ptr,
                                           int dst_pixels_per_line,
                                           unsigned int *sse) {
-  uint16_t FData3[16 * 9];  // Temp data bufffer used in filtering
+  uint16_t fdata3[16 * 9];  // Temp data bufffer used in filtering
   uint8_t temp2[20 * 16];
-  const int16_t *HFilter, *VFilter;
+  const int16_t *hfilter, *vfilter;
 
-  HFilter = VP9_BILINEAR_FILTERS_2TAP(xoffset);
-  VFilter = VP9_BILINEAR_FILTERS_2TAP(yoffset);
+  hfilter = VP9_BILINEAR_FILTERS_2TAP(xoffset);
+  vfilter = VP9_BILINEAR_FILTERS_2TAP(yoffset);
 
-  var_filter_block2d_bil_first_pass(src_ptr, FData3, src_pixels_per_line, 1, 9, 16, HFilter);
-  var_filter_block2d_bil_second_pass(FData3, temp2, 16, 16, 8, 16, VFilter);
+  var_filter_block2d_bil_first_pass(src_ptr, fdata3, src_pixels_per_line,
+                                    1, 9, 16, hfilter);
+  var_filter_block2d_bil_second_pass(fdata3, temp2, 16, 16, 8, 16, vfilter);
 
   return vp9_variance16x8_c(temp2, 16, dst_ptr, dst_pixels_per_line, sse);
 }
@@ -383,16 +526,16 @@ unsigned int vp9_sub_pixel_variance8x16_c(const uint8_t *src_ptr,
                                           const uint8_t *dst_ptr,
                                           int dst_pixels_per_line,
                                           unsigned int *sse) {
-  uint16_t FData3[9 * 16];  // Temp data bufffer used in filtering
+  uint16_t fdata3[9 * 16];  // Temp data bufffer used in filtering
   uint8_t temp2[20 * 16];
-  const int16_t *HFilter, *VFilter;
+  const int16_t *hfilter, *vfilter;
 
-  HFilter = VP9_BILINEAR_FILTERS_2TAP(xoffset);
-  VFilter = VP9_BILINEAR_FILTERS_2TAP(yoffset);
+  hfilter = VP9_BILINEAR_FILTERS_2TAP(xoffset);
+  vfilter = VP9_BILINEAR_FILTERS_2TAP(yoffset);
 
-  var_filter_block2d_bil_first_pass(src_ptr, FData3, src_pixels_per_line,
-                                    1, 17, 8, HFilter);
-  var_filter_block2d_bil_second_pass(FData3, temp2, 8, 8, 16, 8, VFilter);
+  var_filter_block2d_bil_first_pass(src_ptr, fdata3, src_pixels_per_line,
+                                    1, 17, 8, hfilter);
+  var_filter_block2d_bil_second_pass(fdata3, temp2, 8, 8, 16, 8, vfilter);
 
   return vp9_variance8x16_c(temp2, 8, dst_ptr, dst_pixels_per_line, sse);
 }