Merge changes I8a106dd6,Iec442603

* changes:
  d153 intra prediction (16x16) ssse3 using bytes
  d153 intra prediction ssse3 using bytes

diff --combined vp9/common/vp9_rtcd_defs.sh
index 58b7ee7b88d3312ed08cff0adfd89bcf71cf64f4,864a84095cfc614632cc89e837ddde24fc70b754..67dced2102162fdc98e9328785adefe4eadba273
--- 1/vp9/common/vp9_rtcd_defs.sh
--- 2/vp9/common/vp9_rtcd_defs.sh
+++ b/vp9/common/vp9_rtcd_defs.sh
@@@ -27,6 -27,22 +27,6 @@@ forward_decls vp9_common_forward_decl
  # this variable is for functions that are 64 bit only.
  [ $arch = "x86_64" ] && mmx_x86_64=mmx && sse2_x86_64=sse2 && ssse3_x86_64=ssse3
  
 -#
 -# Dequant
 -#
 -
 -prototype void vp9_idct_add_16x16 "int16_t *input, uint8_t *dest, int stride, int eob"
 -specialize vp9_idct_add_16x16
 -
 -prototype void vp9_idct_add_8x8 "int16_t *input, uint8_t *dest, int stride, int eob"
 -specialize vp9_idct_add_8x8
 -
 -prototype void vp9_idct_add "int16_t *input, uint8_t *dest, int stride, int eob"
 -specialize vp9_idct_add
 -
 -prototype void vp9_idct_add_32x32 "int16_t *q, uint8_t *dst, int stride, int eob"
 -specialize vp9_idct_add_32x32
 -
  #
  # RECON
  #
@@@ -49,7 -65,7 +49,7 @@@ prototype void vp9_d135_predictor_4x4 "
  specialize vp9_d135_predictor_4x4
  
  prototype void vp9_d153_predictor_4x4 "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"
- specialize vp9_d153_predictor_4x4
+ specialize vp9_d153_predictor_4x4 $ssse3_x86inc
  
  prototype void vp9_v_predictor_4x4 "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"
  specialize vp9_v_predictor_4x4 $sse_x86inc
@@@ -88,7 -104,7 +88,7 @@@ prototype void vp9_d135_predictor_8x8 "
  specialize vp9_d135_predictor_8x8
  
  prototype void vp9_d153_predictor_8x8 "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"
- specialize vp9_d153_predictor_8x8
+ specialize vp9_d153_predictor_8x8 $ssse3_x86inc
  
  prototype void vp9_v_predictor_8x8 "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"
  specialize vp9_v_predictor_8x8 $sse_x86inc
@@@ -127,7 -143,7 +127,7 @@@ prototype void vp9_d135_predictor_16x1
  specialize vp9_d135_predictor_16x16
  
  prototype void vp9_d153_predictor_16x16 "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"
- specialize vp9_d153_predictor_16x16
+ specialize vp9_d153_predictor_16x16 $ssse3_x86inc
  
  prototype void vp9_v_predictor_16x16 "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"
  specialize vp9_v_predictor_16x16 $sse2_x86inc
@@@ -186,6 -202,17 +186,6 @@@ specialize vp9_dc_left_predictor_32x3
  prototype void vp9_dc_128_predictor_32x32 "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"
  specialize vp9_dc_128_predictor_32x32
  
 -if [ "$CONFIG_VP9_DECODER" = "yes" ]; then
 -prototype void vp9_add_constant_residual_8x8 "const int16_t diff, uint8_t *dest, int stride"
 -specialize vp9_add_constant_residual_8x8 sse2 neon
 -
 -prototype void vp9_add_constant_residual_16x16 "const int16_t diff, uint8_t *dest, int stride"
 -specialize vp9_add_constant_residual_16x16 sse2 neon
 -
 -prototype void vp9_add_constant_residual_32x32 "const int16_t diff, uint8_t *dest, int stride"
 -specialize vp9_add_constant_residual_32x32 sse2 neon
 -fi
 -
  #
  # Loopfilter
  #
@@@ -241,37 -268,37 +241,37 @@@ specialize vp9_blend_
  # Sub Pixel Filters
  #
  prototype void vp9_convolve_copy "const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h"
 -specialize vp9_convolve_copy $sse2_x86inc neon
 +specialize vp9_convolve_copy $sse2_x86inc neon dspr2
  
  prototype void vp9_convolve_avg "const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h"
 -specialize vp9_convolve_avg $sse2_x86inc neon
 +specialize vp9_convolve_avg $sse2_x86inc neon dspr2
  
  prototype void vp9_convolve8 "const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h"
 -specialize vp9_convolve8 ssse3 neon
 +specialize vp9_convolve8 ssse3 neon dspr2
  
  prototype void vp9_convolve8_horiz "const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h"
 -specialize vp9_convolve8_horiz ssse3 neon
 +specialize vp9_convolve8_horiz ssse3 neon dspr2
  
  prototype void vp9_convolve8_vert "const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h"
 -specialize vp9_convolve8_vert ssse3 neon
 +specialize vp9_convolve8_vert ssse3 neon dspr2
  
  prototype void vp9_convolve8_avg "const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h"
 -specialize vp9_convolve8_avg ssse3 neon
 +specialize vp9_convolve8_avg ssse3 neon dspr2
  
  prototype void vp9_convolve8_avg_horiz "const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h"
 -specialize vp9_convolve8_avg_horiz ssse3 neon
 +specialize vp9_convolve8_avg_horiz ssse3 neon dspr2
  
  prototype void vp9_convolve8_avg_vert "const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h"
 -specialize vp9_convolve8_avg_vert ssse3 neon
 +specialize vp9_convolve8_avg_vert ssse3 neon dspr2
  
  #
  # dct
  #
 -prototype void vp9_short_idct4x4_1_add "int16_t *input, uint8_t *dest, int dest_stride"
 -specialize vp9_short_idct4x4_1_add sse2 neon
 +prototype void vp9_idct4x4_1_add "int16_t *input, uint8_t *dest, int dest_stride"
 +specialize vp9_idct4x4_1_add sse2 neon
  
 -prototype void vp9_short_idct4x4_add "int16_t *input, uint8_t *dest, int dest_stride"
 -specialize vp9_short_idct4x4_add sse2 neon
 +prototype void vp9_idct4x4_16_add "int16_t *input, uint8_t *dest, int dest_stride"
 +specialize vp9_idct4x4_16_add sse2 neon
  
  prototype void vp9_short_idct8x8_1_add "int16_t *input, uint8_t *dest, int dest_stride"
  specialize vp9_short_idct8x8_1_add sse2 neon
@@@ -279,8 -306,8 +279,8 @@@
  prototype void vp9_short_idct8x8_add "int16_t *input, uint8_t *dest, int dest_stride"
  specialize vp9_short_idct8x8_add sse2 neon
  
 -prototype void vp9_short_idct10_8x8_add "int16_t *input, uint8_t *dest, int dest_stride"
 -specialize vp9_short_idct10_8x8_add sse2 neon
 +prototype void vp9_short_idct8x8_10_add "int16_t *input, uint8_t *dest, int dest_stride"
 +specialize vp9_short_idct8x8_10_add sse2 neon
  
  prototype void vp9_short_idct16x16_1_add "int16_t *input, uint8_t *dest, int dest_stride"
  specialize vp9_short_idct16x16_1_add sse2 neon
@@@ -288,14 -315,14 +288,14 @@@
  prototype void vp9_short_idct16x16_add "int16_t *input, uint8_t *dest, int dest_stride"
  specialize vp9_short_idct16x16_add sse2 neon
  
 -prototype void vp9_short_idct10_16x16_add "int16_t *input, uint8_t *dest, int dest_stride"
 -specialize vp9_short_idct10_16x16_add sse2 neon
 +prototype void vp9_short_idct16x16_10_add "int16_t *input, uint8_t *dest, int dest_stride"
 +specialize vp9_short_idct16x16_10_add sse2 neon
  
  prototype void vp9_short_idct32x32_add "int16_t *input, uint8_t *dest, int dest_stride"
  specialize vp9_short_idct32x32_add sse2 neon
  
 -prototype void vp9_short_idct1_32x32 "int16_t *input, int16_t *output"
 -specialize vp9_short_idct1_32x32
 +prototype void vp9_short_idct32x32_1_add "int16_t *input, uint8_t *dest, int dest_stride"
 +specialize vp9_short_idct32x32_1_add sse2
  
  prototype void vp9_short_iht4x4_add "int16_t *input, uint8_t *dest, int dest_stride, int tx_type"
  specialize vp9_short_iht4x4_add sse2 neon
@@@ -310,11 -337,11 +310,11 @@@ prototype void vp9_idct4_1d "int16_t *i
  specialize vp9_idct4_1d sse2
  # dct and add
  
 -prototype void vp9_short_iwalsh4x4_1_add "int16_t *input, uint8_t *dest, int dest_stride"
 -specialize vp9_short_iwalsh4x4_1_add
 +prototype void vp9_iwht4x4_1_add "int16_t *input, uint8_t *dest, int dest_stride"
 +specialize vp9_iwht4x4_1_add
  
 -prototype void vp9_short_iwalsh4x4_add "int16_t *input, uint8_t *dest, int dest_stride"
 -specialize vp9_short_iwalsh4x4_add
 +prototype void vp9_iwht4x4_16_add "int16_t *input, uint8_t *dest, int dest_stride"
 +specialize vp9_iwht4x4_16_add
  
  #
  # Encoder functions below this point.