Chromium Code Reviews
chromiumcodereview-hr@appspot.gserviceaccount.com (chromiumcodereview-hr) | Please choose your nickname with Settings | Help | Chromium Project | Gerrit Changes | Sign out
(11)

Unified Diff: source/libvpx/vp9/common/vp9_rtcd_defs.pl

Issue 592203002: libvpx: Pull from upstream (Closed) Base URL: svn://svn.chromium.org/chrome/trunk/deps/third_party/libvpx/
Patch Set: Created 6 years, 3 months ago
Use n/p to move between diff chunks; N/P to move between comments. Draft comments are only viewable by you.
Jump to:
View side-by-side diff with in-line comments
Download patch
« no previous file with comments | « source/libvpx/vp9/common/vp9_reconintra.c ('k') | source/libvpx/vp9/common/vp9_scale.h » ('j') | no next file with comments »
Expand Comments ('e') | Collapse Comments ('c') | Show Comments Hide Comments ('s')
Index: source/libvpx/vp9/common/vp9_rtcd_defs.pl
===================================================================
--- source/libvpx/vp9/common/vp9_rtcd_defs.pl (revision 292072)
+++ source/libvpx/vp9/common/vp9_rtcd_defs.pl (working copy)
@@ -6,6 +6,7 @@
#include "vpx/vpx_integer.h"
#include "vp9/common/vp9_enums.h"
+#include "vp9/common/vp9_idct.h"
struct macroblockd;
@@ -45,6 +46,13 @@
$avx_x86_64 = $avx2_x86_64 = '';
}
+# optimizations which depend on multiple features
+if ((vpx_config("HAVE_AVX2") eq "yes") && (vpx_config("HAVE_SSSE3") eq "yes")) {
+ $avx2_ssse3 = 'avx2';
+} else {
+ $avx2_ssse3 = '';
+}
+
#
# RECON
#
@@ -296,15 +304,15 @@
$vp9_convolve_avg_neon_asm=vp9_convolve_avg_neon;
add_proto qw/void vp9_convolve8/, "const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h";
-specialize qw/vp9_convolve8 sse2 ssse3 neon_asm dspr2 avx2/;
+specialize qw/vp9_convolve8 sse2 ssse3 neon_asm dspr2/, "$avx2_ssse3";
$vp9_convolve8_neon_asm=vp9_convolve8_neon;
add_proto qw/void vp9_convolve8_horiz/, "const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h";
-specialize qw/vp9_convolve8_horiz sse2 ssse3 neon_asm dspr2 avx2/;
+specialize qw/vp9_convolve8_horiz sse2 ssse3 neon_asm dspr2/, "$avx2_ssse3";
$vp9_convolve8_horiz_neon_asm=vp9_convolve8_horiz_neon;
add_proto qw/void vp9_convolve8_vert/, "const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h";
-specialize qw/vp9_convolve8_vert sse2 ssse3 neon_asm dspr2 avx2/;
+specialize qw/vp9_convolve8_vert sse2 ssse3 neon_asm dspr2/, "$avx2_ssse3";
$vp9_convolve8_vert_neon_asm=vp9_convolve8_vert_neon;
add_proto qw/void vp9_convolve8_avg/, "const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h";
@@ -322,69 +330,363 @@
#
# dct
#
-add_proto qw/void vp9_idct4x4_1_add/, "const int16_t *input, uint8_t *dest, int dest_stride";
-specialize qw/vp9_idct4x4_1_add sse2 neon_asm dspr2/;
-$vp9_idct4x4_1_add_neon_asm=vp9_idct4x4_1_add_neon;
+if (vpx_config("CONFIG_VP9_HIGHBITDEPTH") eq "yes") {
+ add_proto qw/void vp9_idct4x4_1_add/, "const tran_low_t *input, uint8_t *dest, int dest_stride";
+ specialize qw/vp9_idct4x4_1_add/;
-add_proto qw/void vp9_idct4x4_16_add/, "const int16_t *input, uint8_t *dest, int dest_stride";
-specialize qw/vp9_idct4x4_16_add sse2 neon_asm dspr2/;
-$vp9_idct4x4_16_add_neon_asm=vp9_idct4x4_16_add_neon;
+ add_proto qw/void vp9_idct4x4_16_add/, "const tran_low_t *input, uint8_t *dest, int dest_stride";
+ specialize qw/vp9_idct4x4_16_add/;
-add_proto qw/void vp9_idct8x8_1_add/, "const int16_t *input, uint8_t *dest, int dest_stride";
-specialize qw/vp9_idct8x8_1_add sse2 neon_asm dspr2/;
-$vp9_idct8x8_1_add_neon_asm=vp9_idct8x8_1_add_neon;
+ add_proto qw/void vp9_idct8x8_1_add/, "const tran_low_t *input, uint8_t *dest, int dest_stride";
+ specialize qw/vp9_idct8x8_1_add/;
-add_proto qw/void vp9_idct8x8_64_add/, "const int16_t *input, uint8_t *dest, int dest_stride";
-specialize qw/vp9_idct8x8_64_add sse2 neon_asm dspr2/, "$ssse3_x86_64";
-$vp9_idct8x8_64_add_neon_asm=vp9_idct8x8_64_add_neon;
+ add_proto qw/void vp9_idct8x8_64_add/, "const tran_low_t *input, uint8_t *dest, int dest_stride";
+ specialize qw/vp9_idct8x8_64_add/;
-add_proto qw/void vp9_idct8x8_12_add/, "const int16_t *input, uint8_t *dest, int dest_stride";
-specialize qw/vp9_idct8x8_12_add sse2 neon_asm dspr2/, "$ssse3_x86_64";
-$vp9_idct8x8_12_add_neon_asm=vp9_idct8x8_12_add_neon;
+ add_proto qw/void vp9_idct8x8_12_add/, "const tran_low_t *input, uint8_t *dest, int dest_stride";
+ specialize qw/vp9_idct8x8_12_add/;
-add_proto qw/void vp9_idct16x16_1_add/, "const int16_t *input, uint8_t *dest, int dest_stride";
-specialize qw/vp9_idct16x16_1_add sse2 neon_asm dspr2/;
-$vp9_idct16x16_1_add_neon_asm=vp9_idct16x16_1_add_neon;
+ add_proto qw/void vp9_idct16x16_1_add/, "const tran_low_t *input, uint8_t *dest, int dest_stride";
+ specialize qw/vp9_idct16x16_1_add/;
-add_proto qw/void vp9_idct16x16_256_add/, "const int16_t *input, uint8_t *dest, int dest_stride";
-specialize qw/vp9_idct16x16_256_add sse2 ssse3 neon_asm dspr2/;
-$vp9_idct16x16_256_add_neon_asm=vp9_idct16x16_256_add_neon;
+ add_proto qw/void vp9_idct16x16_256_add/, "const tran_low_t *input, uint8_t *dest, int dest_stride";
+ specialize qw/vp9_idct16x16_256_add/;
-add_proto qw/void vp9_idct16x16_10_add/, "const int16_t *input, uint8_t *dest, int dest_stride";
-specialize qw/vp9_idct16x16_10_add sse2 ssse3 neon_asm dspr2/;
-$vp9_idct16x16_10_add_neon_asm=vp9_idct16x16_10_add_neon;
+ add_proto qw/void vp9_idct16x16_10_add/, "const tran_low_t *input, uint8_t *dest, int dest_stride";
+ specialize qw/vp9_idct16x16_10_add/;
-add_proto qw/void vp9_idct32x32_1024_add/, "const int16_t *input, uint8_t *dest, int dest_stride";
-specialize qw/vp9_idct32x32_1024_add sse2 neon_asm dspr2/;
-$vp9_idct32x32_1024_add_neon_asm=vp9_idct32x32_1024_add_neon;
+ add_proto qw/void vp9_idct32x32_1024_add/, "const tran_low_t *input, uint8_t *dest, int dest_stride";
+ specialize qw/vp9_idct32x32_1024_add/;
-add_proto qw/void vp9_idct32x32_34_add/, "const int16_t *input, uint8_t *dest, int dest_stride";
-specialize qw/vp9_idct32x32_34_add sse2 neon_asm dspr2/;
-$vp9_idct32x32_34_add_neon_asm=vp9_idct32x32_1024_add_neon;
+ add_proto qw/void vp9_idct32x32_34_add/, "const tran_low_t *input, uint8_t *dest, int dest_stride";
+ specialize qw/vp9_idct32x32_34_add/;
-add_proto qw/void vp9_idct32x32_1_add/, "const int16_t *input, uint8_t *dest, int dest_stride";
-specialize qw/vp9_idct32x32_1_add sse2 neon_asm dspr2/;
-$vp9_idct32x32_1_add_neon_asm=vp9_idct32x32_1_add_neon;
+ add_proto qw/void vp9_idct32x32_1_add/, "const tran_low_t *input, uint8_t *dest, int dest_stride";
+ specialize qw/vp9_idct32x32_1_add/;
-add_proto qw/void vp9_iht4x4_16_add/, "const int16_t *input, uint8_t *dest, int dest_stride, int tx_type";
-specialize qw/vp9_iht4x4_16_add sse2 neon_asm dspr2/;
-$vp9_iht4x4_16_add_neon_asm=vp9_iht4x4_16_add_neon;
+ add_proto qw/void vp9_iht4x4_16_add/, "const tran_low_t *input, uint8_t *dest, int dest_stride, int tx_type";
+ specialize qw/vp9_iht4x4_16_add/;
-add_proto qw/void vp9_iht8x8_64_add/, "const int16_t *input, uint8_t *dest, int dest_stride, int tx_type";
-specialize qw/vp9_iht8x8_64_add sse2 neon_asm dspr2/;
-$vp9_iht8x8_64_add_neon_asm=vp9_iht8x8_64_add_neon;
+ add_proto qw/void vp9_iht8x8_64_add/, "const tran_low_t *input, uint8_t *dest, int dest_stride, int tx_type";
+ specialize qw/vp9_iht8x8_64_add/;
-add_proto qw/void vp9_iht16x16_256_add/, "const int16_t *input, uint8_t *output, int pitch, int tx_type";
-specialize qw/vp9_iht16x16_256_add sse2 dspr2/;
+ add_proto qw/void vp9_iht16x16_256_add/, "const tran_low_t *input, uint8_t *output, int pitch, int tx_type";
+ specialize qw/vp9_iht16x16_256_add/;
-# dct and add
+ # dct and add
-add_proto qw/void vp9_iwht4x4_1_add/, "const int16_t *input, uint8_t *dest, int dest_stride";
-specialize qw/vp9_iwht4x4_1_add/;
+ add_proto qw/void vp9_iwht4x4_1_add/, "const tran_low_t *input, uint8_t *dest, int dest_stride";
+ specialize qw/vp9_iwht4x4_1_add/;
-add_proto qw/void vp9_iwht4x4_16_add/, "const int16_t *input, uint8_t *dest, int dest_stride";
-specialize qw/vp9_iwht4x4_16_add/;
+ add_proto qw/void vp9_iwht4x4_16_add/, "const tran_low_t *input, uint8_t *dest, int dest_stride";
+ specialize qw/vp9_iwht4x4_16_add/;
+} else {
+ add_proto qw/void vp9_idct4x4_1_add/, "const tran_low_t *input, uint8_t *dest, int dest_stride";
+ specialize qw/vp9_idct4x4_1_add sse2 neon_asm dspr2/;
+ $vp9_idct4x4_1_add_neon_asm=vp9_idct4x4_1_add_neon;
+ add_proto qw/void vp9_idct4x4_16_add/, "const tran_low_t *input, uint8_t *dest, int dest_stride";
+ specialize qw/vp9_idct4x4_16_add sse2 neon_asm dspr2/;
+ $vp9_idct4x4_16_add_neon_asm=vp9_idct4x4_16_add_neon;
+
+ add_proto qw/void vp9_idct8x8_1_add/, "const tran_low_t *input, uint8_t *dest, int dest_stride";
+ specialize qw/vp9_idct8x8_1_add sse2 neon_asm dspr2/;
+ $vp9_idct8x8_1_add_neon_asm=vp9_idct8x8_1_add_neon;
+
+ add_proto qw/void vp9_idct8x8_64_add/, "const tran_low_t *input, uint8_t *dest, int dest_stride";
+ specialize qw/vp9_idct8x8_64_add sse2 neon_asm dspr2/, "$ssse3_x86_64";
+ $vp9_idct8x8_64_add_neon_asm=vp9_idct8x8_64_add_neon;
+
+ add_proto qw/void vp9_idct8x8_12_add/, "const tran_low_t *input, uint8_t *dest, int dest_stride";
+ specialize qw/vp9_idct8x8_12_add sse2 neon_asm dspr2/, "$ssse3_x86_64";
+ $vp9_idct8x8_12_add_neon_asm=vp9_idct8x8_12_add_neon;
+
+ add_proto qw/void vp9_idct16x16_1_add/, "const tran_low_t *input, uint8_t *dest, int dest_stride";
+ specialize qw/vp9_idct16x16_1_add sse2 neon_asm dspr2/;
+ $vp9_idct16x16_1_add_neon_asm=vp9_idct16x16_1_add_neon;
+
+ add_proto qw/void vp9_idct16x16_256_add/, "const tran_low_t *input, uint8_t *dest, int dest_stride";
+ specialize qw/vp9_idct16x16_256_add sse2 ssse3 neon_asm dspr2/;
+ $vp9_idct16x16_256_add_neon_asm=vp9_idct16x16_256_add_neon;
+
+ add_proto qw/void vp9_idct16x16_10_add/, "const tran_low_t *input, uint8_t *dest, int dest_stride";
+ specialize qw/vp9_idct16x16_10_add sse2 ssse3 neon_asm dspr2/;
+ $vp9_idct16x16_10_add_neon_asm=vp9_idct16x16_10_add_neon;
+
+ add_proto qw/void vp9_idct32x32_1024_add/, "const tran_low_t *input, uint8_t *dest, int dest_stride";
+ specialize qw/vp9_idct32x32_1024_add sse2 neon_asm dspr2/;
+ $vp9_idct32x32_1024_add_neon_asm=vp9_idct32x32_1024_add_neon;
+
+ add_proto qw/void vp9_idct32x32_34_add/, "const tran_low_t *input, uint8_t *dest, int dest_stride";
+ specialize qw/vp9_idct32x32_34_add sse2 neon_asm dspr2/;
+ $vp9_idct32x32_34_add_neon_asm=vp9_idct32x32_1024_add_neon;
+
+ add_proto qw/void vp9_idct32x32_1_add/, "const tran_low_t *input, uint8_t *dest, int dest_stride";
+ specialize qw/vp9_idct32x32_1_add sse2 neon_asm dspr2/;
+ $vp9_idct32x32_1_add_neon_asm=vp9_idct32x32_1_add_neon;
+
+ add_proto qw/void vp9_iht4x4_16_add/, "const tran_low_t *input, uint8_t *dest, int dest_stride, int tx_type";
+ specialize qw/vp9_iht4x4_16_add sse2 neon_asm dspr2/;
+ $vp9_iht4x4_16_add_neon_asm=vp9_iht4x4_16_add_neon;
+
+ add_proto qw/void vp9_iht8x8_64_add/, "const tran_low_t *input, uint8_t *dest, int dest_stride, int tx_type";
+ specialize qw/vp9_iht8x8_64_add sse2 neon_asm dspr2/;
+ $vp9_iht8x8_64_add_neon_asm=vp9_iht8x8_64_add_neon;
+
+ add_proto qw/void vp9_iht16x16_256_add/, "const tran_low_t *input, uint8_t *output, int pitch, int tx_type";
+ specialize qw/vp9_iht16x16_256_add sse2 dspr2/;
+
+ # dct and add
+
+ add_proto qw/void vp9_iwht4x4_1_add/, "const tran_low_t *input, uint8_t *dest, int dest_stride";
+ specialize qw/vp9_iwht4x4_1_add/;
+
+ add_proto qw/void vp9_iwht4x4_16_add/, "const tran_low_t *input, uint8_t *dest, int dest_stride";
+ specialize qw/vp9_iwht4x4_16_add/;
+}
+
+# High bitdepth functions
+if (vpx_config("CONFIG_VP9_HIGHBITDEPTH") eq "yes") {
+ #
+ # Intra prediction
+ #
+ add_proto qw/void vp9_high_d207_predictor_4x4/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bps";
+ specialize qw/vp9_high_d207_predictor_4x4/;
+
+ add_proto qw/void vp9_high_d45_predictor_4x4/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bps";
+ specialize qw/vp9_high_d45_predictor_4x4/;
+
+ add_proto qw/void vp9_high_d63_predictor_4x4/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bps";
+ specialize qw/vp9_high_d63_predictor_4x4/;
+
+ add_proto qw/void vp9_high_h_predictor_4x4/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bps";
+ specialize qw/vp9_high_h_predictor_4x4/;
+
+ add_proto qw/void vp9_high_d117_predictor_4x4/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bps";
+ specialize qw/vp9_high_d117_predictor_4x4/;
+
+ add_proto qw/void vp9_high_d135_predictor_4x4/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bps";
+ specialize qw/vp9_high_d135_predictor_4x4/;
+
+ add_proto qw/void vp9_high_d153_predictor_4x4/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bps";
+ specialize qw/vp9_high_d153_predictor_4x4/;
+
+ add_proto qw/void vp9_high_v_predictor_4x4/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bps";
+ specialize qw/vp9_high_v_predictor_4x4 neon/, "$sse_x86inc";
+
+ add_proto qw/void vp9_high_tm_predictor_4x4/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bps";
+ specialize qw/vp9_high_tm_predictor_4x4/, "$sse_x86inc";
+
+ add_proto qw/void vp9_high_dc_predictor_4x4/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bps";
+ specialize qw/vp9_high_dc_predictor_4x4/, "$sse_x86inc";
+
+ add_proto qw/void vp9_high_dc_top_predictor_4x4/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bps";
+ specialize qw/vp9_high_dc_top_predictor_4x4/;
+
+ add_proto qw/void vp9_high_dc_left_predictor_4x4/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bps";
+ specialize qw/vp9_high_dc_left_predictor_4x4/;
+
+ add_proto qw/void vp9_high_dc_128_predictor_4x4/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bps";
+ specialize qw/vp9_high_dc_128_predictor_4x4/;
+
+ add_proto qw/void vp9_high_d207_predictor_8x8/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bps";
+ specialize qw/vp9_high_d207_predictor_8x8/;
+
+ add_proto qw/void vp9_high_d45_predictor_8x8/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bps";
+ specialize qw/vp9_high_d45_predictor_8x8/;
+
+ add_proto qw/void vp9_high_d63_predictor_8x8/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bps";
+ specialize qw/vp9_high_d63_predictor_8x8/;
+
+ add_proto qw/void vp9_high_h_predictor_8x8/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bps";
+ specialize qw/vp9_high_h_predictor_8x8/;
+
+ add_proto qw/void vp9_high_d117_predictor_8x8/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bps";
+ specialize qw/vp9_high_d117_predictor_8x8/;
+
+ add_proto qw/void vp9_high_d135_predictor_8x8/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bps";
+ specialize qw/vp9_high_d135_predictor_8x8/;
+
+ add_proto qw/void vp9_high_d153_predictor_8x8/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bps";
+ specialize qw/vp9_high_d153_predictor_8x8/;
+
+ add_proto qw/void vp9_high_v_predictor_8x8/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bps";
+ specialize qw/vp9_high_v_predictor_8x8/, "$sse2_x86inc";
+
+ add_proto qw/void vp9_high_tm_predictor_8x8/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bps";
+ specialize qw/vp9_high_tm_predictor_8x8/, "$sse2_x86inc";
+
+ add_proto qw/void vp9_high_dc_predictor_8x8/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bps";
+ specialize qw/vp9_high_dc_predictor_8x8/, "$sse2_x86inc";;
+
+ add_proto qw/void vp9_high_dc_top_predictor_8x8/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bps";
+ specialize qw/vp9_high_dc_top_predictor_8x8/;
+
+ add_proto qw/void vp9_high_dc_left_predictor_8x8/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bps";
+ specialize qw/vp9_high_dc_left_predictor_8x8/;
+
+ add_proto qw/void vp9_high_dc_128_predictor_8x8/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bps";
+ specialize qw/vp9_high_dc_128_predictor_8x8/;
+
+ add_proto qw/void vp9_high_d207_predictor_16x16/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bps";
+ specialize qw/vp9_high_d207_predictor_16x16/;
+
+ add_proto qw/void vp9_high_d45_predictor_16x16/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bps";
+ specialize qw/vp9_high_d45_predictor_16x16/;
+
+ add_proto qw/void vp9_high_d63_predictor_16x16/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bps";
+ specialize qw/vp9_high_d63_predictor_16x16/;
+
+ add_proto qw/void vp9_high_h_predictor_16x16/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bps";
+ specialize qw/vp9_high_h_predictor_16x16/;
+
+ add_proto qw/void vp9_high_d117_predictor_16x16/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bps";
+ specialize qw/vp9_high_d117_predictor_16x16/;
+
+ add_proto qw/void vp9_high_d135_predictor_16x16/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bps";
+ specialize qw/vp9_high_d135_predictor_16x16/;
+
+ add_proto qw/void vp9_high_d153_predictor_16x16/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bps";
+ specialize qw/vp9_high_d153_predictor_16x16/;
+
+ add_proto qw/void vp9_high_v_predictor_16x16/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bps";
+ specialize qw/vp9_high_v_predictor_16x16 neon/, "$sse2_x86inc";
+
+ add_proto qw/void vp9_high_tm_predictor_16x16/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bps";
+ specialize qw/vp9_high_tm_predictor_16x16/, "$sse2_x86_64";
+
+ add_proto qw/void vp9_high_dc_predictor_16x16/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bps";
+ specialize qw/vp9_high_dc_predictor_16x16/, "$sse2_x86inc";
+
+ add_proto qw/void vp9_high_dc_top_predictor_16x16/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bps";
+ specialize qw/vp9_high_dc_top_predictor_16x16/;
+
+ add_proto qw/void vp9_high_dc_left_predictor_16x16/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bps";
+ specialize qw/vp9_high_dc_left_predictor_16x16/;
+
+ add_proto qw/void vp9_high_dc_128_predictor_16x16/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bps";
+ specialize qw/vp9_high_dc_128_predictor_16x16/;
+
+ add_proto qw/void vp9_high_d207_predictor_32x32/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bps";
+ specialize qw/vp9_high_d207_predictor_32x32/;
+
+ add_proto qw/void vp9_high_d45_predictor_32x32/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bps";
+ specialize qw/vp9_high_d45_predictor_32x32/;
+
+ add_proto qw/void vp9_high_d63_predictor_32x32/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bps";
+ specialize qw/vp9_high_d63_predictor_32x32/;
+
+ add_proto qw/void vp9_high_h_predictor_32x32/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bps";
+ specialize qw/vp9_high_h_predictor_32x32/;
+
+ add_proto qw/void vp9_high_d117_predictor_32x32/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bps";
+ specialize qw/vp9_high_d117_predictor_32x32/;
+
+ add_proto qw/void vp9_high_d135_predictor_32x32/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bps";
+ specialize qw/vp9_high_d135_predictor_32x32/;
+
+ add_proto qw/void vp9_high_d153_predictor_32x32/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bps";
+ specialize qw/vp9_high_d153_predictor_32x32/;
+
+ add_proto qw/void vp9_high_v_predictor_32x32/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bps";
+ specialize qw/vp9_high_v_predictor_32x32/, "$sse2_x86inc";
+
+ add_proto qw/void vp9_high_tm_predictor_32x32/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bps";
+ specialize qw/vp9_high_tm_predictor_32x32/, "$sse2_x86_64";
+
+ add_proto qw/void vp9_high_dc_predictor_32x32/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bps";
+ specialize qw/vp9_high_dc_predictor_32x32/, "$sse2_x86_64";
+
+ add_proto qw/void vp9_high_dc_top_predictor_32x32/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bps";
+ specialize qw/vp9_high_dc_top_predictor_32x32/;
+
+ add_proto qw/void vp9_high_dc_left_predictor_32x32/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bps";
+ specialize qw/vp9_high_dc_left_predictor_32x32/;
+
+ add_proto qw/void vp9_high_dc_128_predictor_32x32/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bps";
+ specialize qw/vp9_high_dc_128_predictor_32x32/;
+
+ #
+ # Sub Pixel Filters
+ #
+ add_proto qw/void vp9_high_convolve_copy/, "const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps";
+ specialize qw/vp9_high_convolve_copy/;
+
+ add_proto qw/void vp9_high_convolve_avg/, "const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps";
+ specialize qw/vp9_high_convolve_avg/;
+
+ add_proto qw/void vp9_high_convolve8/, "const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps";
+ specialize qw/vp9_high_convolve8/, "$sse2_x86_64";
+
+ add_proto qw/void vp9_high_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, int bps";
+ specialize qw/vp9_high_convolve8_horiz/, "$sse2_x86_64";
+
+ add_proto qw/void vp9_high_convolve8_vert/, "const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps";
+ specialize qw/vp9_high_convolve8_vert/, "$sse2_x86_64";
+
+ add_proto qw/void vp9_high_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, int bps";
+ specialize qw/vp9_high_convolve8_avg/, "$sse2_x86_64";
+
+ add_proto qw/void vp9_high_convolve8_avg_horiz/, "const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps";
+ specialize qw/vp9_high_convolve8_avg_horiz/, "$sse2_x86_64";
+
+ add_proto qw/void vp9_high_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, int bps";
+ specialize qw/vp9_high_convolve8_avg_vert/, "$sse2_x86_64";
+
+ #
+ # dct
+ #
+ add_proto qw/void vp9_high_idct4x4_1_add/, "const tran_low_t *input, uint8_t *dest, int dest_stride, int bd";
+ specialize qw/vp9_high_idct4x4_1_add/;
+
+ add_proto qw/void vp9_high_idct4x4_16_add/, "const tran_low_t *input, uint8_t *dest, int dest_stride, int bd";
+ specialize qw/vp9_high_idct4x4_16_add/;
+
+ add_proto qw/void vp9_high_idct8x8_1_add/, "const tran_low_t *input, uint8_t *dest, int dest_stride, int bd";
+ specialize qw/vp9_high_idct8x8_1_add/;
+
+ add_proto qw/void vp9_high_idct8x8_64_add/, "const tran_low_t *input, uint8_t *dest, int dest_stride, int bd";
+ specialize qw/vp9_high_idct8x8_64_add/;
+
+ add_proto qw/void vp9_high_idct8x8_10_add/, "const tran_low_t *input, uint8_t *dest, int dest_stride, int bd";
+ specialize qw/vp9_high_idct8x8_10_add/;
+
+ add_proto qw/void vp9_high_idct16x16_1_add/, "const tran_low_t *input, uint8_t *dest, int dest_stride, int bd";
+ specialize qw/vp9_high_idct16x16_1_add/;
+
+ add_proto qw/void vp9_high_idct16x16_256_add/, "const tran_low_t *input, uint8_t *dest, int dest_stride, int bd";
+ specialize qw/vp9_high_idct16x16_256_add/;
+
+ add_proto qw/void vp9_high_idct16x16_10_add/, "const tran_low_t *input, uint8_t *dest, int dest_stride, int bd";
+ specialize qw/vp9_high_idct16x16_10_add/;
+
+ add_proto qw/void vp9_high_idct32x32_1024_add/, "const tran_low_t *input, uint8_t *dest, int dest_stride, int bd";
+ specialize qw/vp9_high_idct32x32_1024_add/;
+
+ add_proto qw/void vp9_high_idct32x32_34_add/, "const tran_low_t *input, uint8_t *dest, int dest_stride, int bd";
+ specialize qw/vp9_high_idct32x32_34_add/;
+
+ add_proto qw/void vp9_high_idct32x32_1_add/, "const tran_low_t *input, uint8_t *dest, int dest_stride, int bd";
+ specialize qw/vp9_high_idct32x32_1_add/;
+
+ add_proto qw/void vp9_high_iht4x4_16_add/, "const tran_low_t *input, uint8_t *dest, int dest_stride, int tx_type, int bd";
+ specialize qw/vp9_high_iht4x4_16_add/;
+
+ add_proto qw/void vp9_high_iht8x8_64_add/, "const tran_low_t *input, uint8_t *dest, int dest_stride, int tx_type, int bd";
+ specialize qw/vp9_high_iht8x8_64_add/;
+
+ add_proto qw/void vp9_high_iht16x16_256_add/, "const tran_low_t *input, uint8_t *output, int pitch, int tx_type, int bd";
+ specialize qw/vp9_high_iht16x16_256_add/;
+
+ # dct and add
+
+ add_proto qw/void vp9_high_iwht4x4_1_add/, "const tran_low_t *input, uint8_t *dest, int dest_stride, int bd";
+ specialize qw/vp9_high_iwht4x4_1_add/;
+
+ add_proto qw/void vp9_high_iwht4x4_16_add/, "const tran_low_t *input, uint8_t *dest, int dest_stride, int bd";
+ specialize qw/vp9_high_iwht4x4_16_add/;
+}
+
#
# Encoder functions below this point.
#
@@ -699,24 +1001,43 @@
specialize qw/vp9_get_mb_ss/, "$sse2_x86inc";
# ENCODEMB INVOKE
-add_proto qw/int64_t vp9_block_error/, "const int16_t *coeff, const int16_t *dqcoeff, intptr_t block_size, int64_t *ssz";
-specialize qw/vp9_block_error avx2/, "$sse2_x86inc";
-
add_proto qw/void vp9_subtract_block/, "int rows, int cols, int16_t *diff_ptr, ptrdiff_t diff_stride, const uint8_t *src_ptr, ptrdiff_t src_stride, const uint8_t *pred_ptr, ptrdiff_t pred_stride";
specialize qw/vp9_subtract_block neon/, "$sse2_x86inc";
-add_proto qw/void vp9_quantize_fp/, "const int16_t *coeff_ptr, intptr_t n_coeffs, int skip_block, const int16_t *zbin_ptr, const int16_t *round_ptr, const int16_t *quant_ptr, const int16_t *quant_shift_ptr, int16_t *qcoeff_ptr, int16_t *dqcoeff_ptr, const int16_t *dequant_ptr, int zbin_oq_value, uint16_t *eob_ptr, const int16_t *scan, const int16_t *iscan";
-specialize qw/vp9_quantize_fp neon/, "$ssse3_x86_64";
+if (vpx_config("CONFIG_VP9_HIGHBITDEPTH") eq "yes") {
+# the transform coefficients are held in 32-bit
+# values, so the assembler code for vp9_block_error can no longer be used.
+ add_proto qw/int64_t vp9_block_error/, "const tran_low_t *coeff, const tran_low_t *dqcoeff, intptr_t block_size, int64_t *ssz";
+ specialize qw/vp9_block_error/;
-add_proto qw/void vp9_quantize_fp_32x32/, "const int16_t *coeff_ptr, intptr_t n_coeffs, int skip_block, const int16_t *zbin_ptr, const int16_t *round_ptr, const int16_t *quant_ptr, const int16_t *quant_shift_ptr, int16_t *qcoeff_ptr, int16_t *dqcoeff_ptr, const int16_t *dequant_ptr, int zbin_oq_value, uint16_t *eob_ptr, const int16_t *scan, const int16_t *iscan";
-specialize qw/vp9_quantize_fp_32x32/, "$ssse3_x86_64";
+ add_proto qw/void vp9_quantize_fp/, "const tran_low_t *coeff_ptr, intptr_t n_coeffs, int skip_block, const int16_t *zbin_ptr, const int16_t *round_ptr, const int16_t *quant_ptr, const int16_t *quant_shift_ptr, tran_low_t *qcoeff_ptr, tran_low_t *dqcoeff_ptr, const int16_t *dequant_ptr, int zbin_oq_value, uint16_t *eob_ptr, const int16_t *scan, const int16_t *iscan";
+ specialize qw/vp9_quantize_fp/;
-add_proto qw/void vp9_quantize_b/, "const int16_t *coeff_ptr, intptr_t n_coeffs, int skip_block, const int16_t *zbin_ptr, const int16_t *round_ptr, const int16_t *quant_ptr, const int16_t *quant_shift_ptr, int16_t *qcoeff_ptr, int16_t *dqcoeff_ptr, const int16_t *dequant_ptr, int zbin_oq_value, uint16_t *eob_ptr, const int16_t *scan, const int16_t *iscan";
-specialize qw/vp9_quantize_b/, "$ssse3_x86_64";
+ add_proto qw/void vp9_quantize_fp_32x32/, "const tran_low_t *coeff_ptr, intptr_t n_coeffs, int skip_block, const int16_t *zbin_ptr, const int16_t *round_ptr, const int16_t *quant_ptr, const int16_t *quant_shift_ptr, tran_low_t *qcoeff_ptr, tran_low_t *dqcoeff_ptr, const int16_t *dequant_ptr, int zbin_oq_value, uint16_t *eob_ptr, const int16_t *scan, const int16_t *iscan";
+ specialize qw/vp9_quantize_fp_32x32/;
-add_proto qw/void vp9_quantize_b_32x32/, "const int16_t *coeff_ptr, intptr_t n_coeffs, int skip_block, const int16_t *zbin_ptr, const int16_t *round_ptr, const int16_t *quant_ptr, const int16_t *quant_shift_ptr, int16_t *qcoeff_ptr, int16_t *dqcoeff_ptr, const int16_t *dequant_ptr, int zbin_oq_value, uint16_t *eob_ptr, const int16_t *scan, const int16_t *iscan";
-specialize qw/vp9_quantize_b_32x32/, "$ssse3_x86_64";
+ add_proto qw/void vp9_quantize_b/, "const tran_low_t *coeff_ptr, intptr_t n_coeffs, int skip_block, const int16_t *zbin_ptr, const int16_t *round_ptr, const int16_t *quant_ptr, const int16_t *quant_shift_ptr, tran_low_t *qcoeff_ptr, tran_low_t *dqcoeff_ptr, const int16_t *dequant_ptr, int zbin_oq_value, uint16_t *eob_ptr, const int16_t *scan, const int16_t *iscan";
+ specialize qw/vp9_quantize_b/;
+ add_proto qw/void vp9_quantize_b_32x32/, "const tran_low_t *coeff_ptr, intptr_t n_coeffs, int skip_block, const int16_t *zbin_ptr, const int16_t *round_ptr, const int16_t *quant_ptr, const int16_t *quant_shift_ptr, tran_low_t *qcoeff_ptr, tran_low_t *dqcoeff_ptr, const int16_t *dequant_ptr, int zbin_oq_value, uint16_t *eob_ptr, const int16_t *scan, const int16_t *iscan";
+ specialize qw/vp9_quantize_b_32x32/;
+} else {
+ add_proto qw/int64_t vp9_block_error/, "const tran_low_t *coeff, const tran_low_t *dqcoeff, intptr_t block_size, int64_t *ssz";
+ specialize qw/vp9_block_error avx2/, "$sse2_x86inc";
+
+ add_proto qw/void vp9_quantize_fp/, "const tran_low_t *coeff_ptr, intptr_t n_coeffs, int skip_block, const int16_t *zbin_ptr, const int16_t *round_ptr, const int16_t *quant_ptr, const int16_t *quant_shift_ptr, tran_low_t *qcoeff_ptr, tran_low_t *dqcoeff_ptr, const int16_t *dequant_ptr, int zbin_oq_value, uint16_t *eob_ptr, const int16_t *scan, const int16_t *iscan";
+ specialize qw/vp9_quantize_fp neon/, "$ssse3_x86_64";
+
+ add_proto qw/void vp9_quantize_fp_32x32/, "const tran_low_t *coeff_ptr, intptr_t n_coeffs, int skip_block, const int16_t *zbin_ptr, const int16_t *round_ptr, const int16_t *quant_ptr, const int16_t *quant_shift_ptr, tran_low_t *qcoeff_ptr, tran_low_t *dqcoeff_ptr, const int16_t *dequant_ptr, int zbin_oq_value, uint16_t *eob_ptr, const int16_t *scan, const int16_t *iscan";
+ specialize qw/vp9_quantize_fp_32x32/, "$ssse3_x86_64";
+
+ add_proto qw/void vp9_quantize_b/, "const tran_low_t *coeff_ptr, intptr_t n_coeffs, int skip_block, const int16_t *zbin_ptr, const int16_t *round_ptr, const int16_t *quant_ptr, const int16_t *quant_shift_ptr, tran_low_t *qcoeff_ptr, tran_low_t *dqcoeff_ptr, const int16_t *dequant_ptr, int zbin_oq_value, uint16_t *eob_ptr, const int16_t *scan, const int16_t *iscan";
+ specialize qw/vp9_quantize_b/, "$ssse3_x86_64";
+
+ add_proto qw/void vp9_quantize_b_32x32/, "const tran_low_t *coeff_ptr, intptr_t n_coeffs, int skip_block, const int16_t *zbin_ptr, const int16_t *round_ptr, const int16_t *quant_ptr, const int16_t *quant_shift_ptr, tran_low_t *qcoeff_ptr, tran_low_t *dqcoeff_ptr, const int16_t *dequant_ptr, int zbin_oq_value, uint16_t *eob_ptr, const int16_t *scan, const int16_t *iscan";
+ specialize qw/vp9_quantize_b_32x32/, "$ssse3_x86_64";
+}
+
#
# Structured Similarity (SSIM)
#
@@ -729,45 +1050,87 @@
}
# fdct functions
-add_proto qw/void vp9_fht4x4/, "const int16_t *input, int16_t *output, int stride, int tx_type";
-specialize qw/vp9_fht4x4 sse2/;
-add_proto qw/void vp9_fht8x8/, "const int16_t *input, int16_t *output, int stride, int tx_type";
-specialize qw/vp9_fht8x8 sse2/;
+if (vpx_config("CONFIG_VP9_HIGHBITDEPTH") eq "yes") {
+ add_proto qw/void vp9_fht4x4/, "const int16_t *input, tran_low_t *output, int stride, int tx_type";
+ specialize qw/vp9_fht4x4/;
-add_proto qw/void vp9_fht16x16/, "const int16_t *input, int16_t *output, int stride, int tx_type";
-specialize qw/vp9_fht16x16 sse2/;
+ add_proto qw/void vp9_fht8x8/, "const int16_t *input, tran_low_t *output, int stride, int tx_type";
+ specialize qw/vp9_fht8x8/;
-add_proto qw/void vp9_fwht4x4/, "const int16_t *input, int16_t *output, int stride";
-specialize qw/vp9_fwht4x4/, "$mmx_x86inc";
+ add_proto qw/void vp9_fht16x16/, "const int16_t *input, tran_low_t *output, int stride, int tx_type";
+ specialize qw/vp9_fht16x16/;
-add_proto qw/void vp9_fdct4x4_1/, "const int16_t *input, int16_t *output, int stride";
-specialize qw/vp9_fdct4x4_1 sse2/;
+ add_proto qw/void vp9_fwht4x4/, "const int16_t *input, tran_low_t *output, int stride";
+ specialize qw/vp9_fwht4x4/;
-add_proto qw/void vp9_fdct4x4/, "const int16_t *input, int16_t *output, int stride";
-specialize qw/vp9_fdct4x4 sse2/;
+ add_proto qw/void vp9_fdct4x4_1/, "const int16_t *input, tran_low_t *output, int stride";
+ specialize qw/vp9_fdct4x4_1/;
-add_proto qw/void vp9_fdct8x8_1/, "const int16_t *input, int16_t *output, int stride";
-specialize qw/vp9_fdct8x8_1 sse2 neon/;
+ add_proto qw/void vp9_fdct4x4/, "const int16_t *input, tran_low_t *output, int stride";
+ specialize qw/vp9_fdct4x4/;
-add_proto qw/void vp9_fdct8x8/, "const int16_t *input, int16_t *output, int stride";
-specialize qw/vp9_fdct8x8 sse2 neon/, "$ssse3_x86_64";
+ add_proto qw/void vp9_fdct8x8_1/, "const int16_t *input, tran_low_t *output, int stride";
+ specialize qw/vp9_fdct8x8_1/;
-add_proto qw/void vp9_fdct16x16_1/, "const int16_t *input, int16_t *output, int stride";
-specialize qw/vp9_fdct16x16_1 sse2/;
+ add_proto qw/void vp9_fdct8x8/, "const int16_t *input, tran_low_t *output, int stride";
+ specialize qw/vp9_fdct8x8/;
-add_proto qw/void vp9_fdct16x16/, "const int16_t *input, int16_t *output, int stride";
-specialize qw/vp9_fdct16x16 sse2/;
+ add_proto qw/void vp9_fdct16x16_1/, "const int16_t *input, tran_low_t *output, int stride";
+ specialize qw/vp9_fdct16x16_1/;
-add_proto qw/void vp9_fdct32x32_1/, "const int16_t *input, int16_t *output, int stride";
-specialize qw/vp9_fdct32x32_1 sse2/;
+ add_proto qw/void vp9_fdct16x16/, "const int16_t *input, tran_low_t *output, int stride";
+ specialize qw/vp9_fdct16x16/;
-add_proto qw/void vp9_fdct32x32/, "const int16_t *input, int16_t *output, int stride";
-specialize qw/vp9_fdct32x32 sse2 avx2/;
+ add_proto qw/void vp9_fdct32x32_1/, "const int16_t *input, tran_low_t *output, int stride";
+ specialize qw/vp9_fdct32x32_1/;
-add_proto qw/void vp9_fdct32x32_rd/, "const int16_t *input, int16_t *output, int stride";
-specialize qw/vp9_fdct32x32_rd sse2 avx2/;
+ add_proto qw/void vp9_fdct32x32/, "const int16_t *input, tran_low_t *output, int stride";
+ specialize qw/vp9_fdct32x32/;
+ add_proto qw/void vp9_fdct32x32_rd/, "const int16_t *input, tran_low_t *output, int stride";
+ specialize qw/vp9_fdct32x32_rd/;
+} else {
+ add_proto qw/void vp9_fht4x4/, "const int16_t *input, tran_low_t *output, int stride, int tx_type";
+ specialize qw/vp9_fht4x4 sse2/;
+
+ add_proto qw/void vp9_fht8x8/, "const int16_t *input, tran_low_t *output, int stride, int tx_type";
+ specialize qw/vp9_fht8x8 sse2/;
+
+ add_proto qw/void vp9_fht16x16/, "const int16_t *input, tran_low_t *output, int stride, int tx_type";
+ specialize qw/vp9_fht16x16 sse2/;
+
+ add_proto qw/void vp9_fwht4x4/, "const int16_t *input, tran_low_t *output, int stride";
+ specialize qw/vp9_fwht4x4/, "$mmx_x86inc";
+
+ add_proto qw/void vp9_fdct4x4_1/, "const int16_t *input, tran_low_t *output, int stride";
+ specialize qw/vp9_fdct4x4_1 sse2/;
+
+ add_proto qw/void vp9_fdct4x4/, "const int16_t *input, tran_low_t *output, int stride";
+ specialize qw/vp9_fdct4x4 sse2/;
+
+ add_proto qw/void vp9_fdct8x8_1/, "const int16_t *input, tran_low_t *output, int stride";
+ specialize qw/vp9_fdct8x8_1 sse2 neon/;
+
+ add_proto qw/void vp9_fdct8x8/, "const int16_t *input, tran_low_t *output, int stride";
+ specialize qw/vp9_fdct8x8 sse2 neon/, "$ssse3_x86_64";
+
+ add_proto qw/void vp9_fdct16x16_1/, "const int16_t *input, tran_low_t *output, int stride";
+ specialize qw/vp9_fdct16x16_1 sse2/;
+
+ add_proto qw/void vp9_fdct16x16/, "const int16_t *input, tran_low_t *output, int stride";
+ specialize qw/vp9_fdct16x16 sse2/;
+
+ add_proto qw/void vp9_fdct32x32_1/, "const int16_t *input, tran_low_t *output, int stride";
+ specialize qw/vp9_fdct32x32_1 sse2/;
+
+ add_proto qw/void vp9_fdct32x32/, "const int16_t *input, tran_low_t *output, int stride";
+ specialize qw/vp9_fdct32x32 sse2 avx2/;
+
+ add_proto qw/void vp9_fdct32x32_rd/, "const int16_t *input, tran_low_t *output, int stride";
+ specialize qw/vp9_fdct32x32_rd sse2 avx2/;
+}
+
#
# Motion search
#
@@ -788,6 +1151,654 @@
add_proto qw/void vp9_temporal_filter_apply/, "uint8_t *frame1, unsigned int stride, uint8_t *frame2, unsigned int block_width, unsigned int block_height, int strength, int filter_weight, unsigned int *accumulator, uint16_t *count";
specialize qw/vp9_temporal_filter_apply sse2/;
+if (vpx_config("CONFIG_VP9_HIGHBITDEPTH") eq "yes") {
+
+ # variance
+ add_proto qw/unsigned int vp9_high_variance32x16/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
+ specialize qw/vp9_high_variance32x16/;
+
+ add_proto qw/unsigned int vp9_high_variance16x32/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
+ specialize qw/vp9_high_variance16x32/;
+
+ add_proto qw/unsigned int vp9_high_variance64x32/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
+ specialize qw/vp9_high_variance64x32/;
+
+ add_proto qw/unsigned int vp9_high_variance32x64/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
+ specialize qw/vp9_high_variance32x64/;
+
+ add_proto qw/unsigned int vp9_high_variance32x32/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
+ specialize qw/vp9_high_variance32x32/;
+
+ add_proto qw/unsigned int vp9_high_variance64x64/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
+ specialize qw/vp9_high_variance64x64/;
+
+ add_proto qw/unsigned int vp9_high_variance16x16/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
+ specialize qw/vp9_high_variance16x16/;
+
+ add_proto qw/unsigned int vp9_high_variance16x8/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
+ specialize qw/vp9_high_variance16x8/;
+
+ add_proto qw/unsigned int vp9_high_variance8x16/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
+ specialize qw/vp9_high_variance8x16/;
+
+ add_proto qw/unsigned int vp9_high_variance8x8/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
+ specialize qw/vp9_high_variance8x8/;
+
+ add_proto qw/unsigned int vp9_high_variance8x4/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
+ specialize qw/vp9_high_variance8x4/;
+
+ add_proto qw/unsigned int vp9_high_variance4x8/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
+ specialize qw/vp9_high_variance4x8/;
+
+ add_proto qw/unsigned int vp9_high_variance4x4/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
+ specialize qw/vp9_high_variance4x4/;
+
+ add_proto qw/void vp9_high_get8x8var/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, int *sum";
+ specialize qw/vp9_high_get8x8var/;
+
+ add_proto qw/void vp9_high_get16x16var/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, int *sum";
+ specialize qw/vp9_high_get16x16var/;
+
+ add_proto qw/unsigned int vp9_high_10_variance32x16/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
+ specialize qw/vp9_high_10_variance32x16/;
+
+ add_proto qw/unsigned int vp9_high_10_variance16x32/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
+ specialize qw/vp9_high_10_variance16x32/;
+
+ add_proto qw/unsigned int vp9_high_10_variance64x32/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
+ specialize qw/vp9_high_10_variance64x32/;
+
+ add_proto qw/unsigned int vp9_high_10_variance32x64/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
+ specialize qw/vp9_high_10_variance32x64/;
+
+ add_proto qw/unsigned int vp9_high_10_variance32x32/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
+ specialize qw/vp9_high_10_variance32x32/;
+
+ add_proto qw/unsigned int vp9_high_10_variance64x64/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
+ specialize qw/vp9_high_10_variance64x64/;
+
+ add_proto qw/unsigned int vp9_high_10_variance16x16/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
+ specialize qw/vp9_high_10_variance16x16/;
+
+ add_proto qw/unsigned int vp9_high_10_variance16x8/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
+ specialize qw/vp9_high_10_variance16x8/;
+
+ add_proto qw/unsigned int vp9_high_10_variance8x16/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
+ specialize qw/vp9_high_10_variance8x16/;
+
+ add_proto qw/unsigned int vp9_high_10_variance8x8/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
+ specialize qw/vp9_high_10_variance8x8/;
+
+ add_proto qw/unsigned int vp9_high_10_variance8x4/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
+ specialize qw/vp9_high_10_variance8x4/;
+
+ add_proto qw/unsigned int vp9_high_10_variance4x8/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
+ specialize qw/vp9_high_10_variance4x8/;
+
+ add_proto qw/unsigned int vp9_high_10_variance4x4/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
+ specialize qw/vp9_high_10_variance4x4/;
+
+ add_proto qw/void vp9_high_10_get8x8var/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, int *sum";
+ specialize qw/vp9_high_10_get8x8var/;
+
+ add_proto qw/void vp9_high_10_get16x16var/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, int *sum";
+ specialize qw/vp9_high_10_get16x16var/;
+
+ add_proto qw/unsigned int vp9_high_12_variance32x16/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
+ specialize qw/vp9_high_12_variance32x16/;
+
+ add_proto qw/unsigned int vp9_high_12_variance16x32/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
+ specialize qw/vp9_high_12_variance16x32/;
+
+ add_proto qw/unsigned int vp9_high_12_variance64x32/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
+ specialize qw/vp9_high_12_variance64x32/;
+
+ add_proto qw/unsigned int vp9_high_12_variance32x64/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
+ specialize qw/vp9_high_12_variance32x64/;
+
+ add_proto qw/unsigned int vp9_high_12_variance32x32/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
+ specialize qw/vp9_high_12_variance32x32/;
+
+ add_proto qw/unsigned int vp9_high_12_variance64x64/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
+ specialize qw/vp9_high_12_variance64x64/;
+
+ add_proto qw/unsigned int vp9_high_12_variance16x16/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
+ specialize qw/vp9_high_12_variance16x16/;
+
+ add_proto qw/unsigned int vp9_high_12_variance16x8/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
+ specialize qw/vp9_high_12_variance16x8/;
+
+ add_proto qw/unsigned int vp9_high_12_variance8x16/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
+ specialize qw/vp9_high_12_variance8x16/;
+
+ add_proto qw/unsigned int vp9_high_12_variance8x8/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
+ specialize qw/vp9_high_12_variance8x8/;
+
+ add_proto qw/unsigned int vp9_high_12_variance8x4/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
+ specialize qw/vp9_high_12_variance8x4/;
+
+ add_proto qw/unsigned int vp9_high_12_variance4x8/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
+ specialize qw/vp9_high_12_variance4x8/;
+
+ add_proto qw/unsigned int vp9_high_12_variance4x4/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
+ specialize qw/vp9_high_12_variance4x4/;
+
+ add_proto qw/void vp9_high_12_get8x8var/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, int *sum";
+ specialize qw/vp9_high_12_get8x8var/;
+
+ add_proto qw/void vp9_high_12_get16x16var/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, int *sum";
+ specialize qw/vp9_high_12_get16x16var/;
+
+ add_proto qw/unsigned int vp9_high_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 qw/vp9_high_sub_pixel_variance64x64/;
+
+ add_proto qw/unsigned int vp9_high_sub_pixel_avg_variance64x64/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, const uint8_t *second_pred";
+ specialize qw/vp9_high_sub_pixel_avg_variance64x64/;
+
+ add_proto qw/unsigned int vp9_high_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 qw/vp9_high_sub_pixel_variance32x64/;
+
+ add_proto qw/unsigned int vp9_high_sub_pixel_avg_variance32x64/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, const uint8_t *second_pred";
+ specialize qw/vp9_high_sub_pixel_avg_variance32x64/;
+
+ add_proto qw/unsigned int vp9_high_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 qw/vp9_high_sub_pixel_variance64x32/;
+
+ add_proto qw/unsigned int vp9_high_sub_pixel_avg_variance64x32/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, const uint8_t *second_pred";
+ specialize qw/vp9_high_sub_pixel_avg_variance64x32/;
+
+ add_proto qw/unsigned int vp9_high_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 qw/vp9_high_sub_pixel_variance32x16/;
+
+ add_proto qw/unsigned int vp9_high_sub_pixel_avg_variance32x16/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, const uint8_t *second_pred";
+ specialize qw/vp9_high_sub_pixel_avg_variance32x16/;
+
+ add_proto qw/unsigned int vp9_high_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 qw/vp9_high_sub_pixel_variance16x32/;
+
+ add_proto qw/unsigned int vp9_high_sub_pixel_avg_variance16x32/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, const uint8_t *second_pred";
+ specialize qw/vp9_high_sub_pixel_avg_variance16x32/;
+
+ add_proto qw/unsigned int vp9_high_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 qw/vp9_high_sub_pixel_variance32x32/;
+
+ add_proto qw/unsigned int vp9_high_sub_pixel_avg_variance32x32/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, const uint8_t *second_pred";
+ specialize qw/vp9_high_sub_pixel_avg_variance32x32/;
+
+ add_proto qw/unsigned int vp9_high_sub_pixel_variance16x16/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
+ specialize qw/vp9_high_sub_pixel_variance16x16/;
+
+ add_proto qw/unsigned int vp9_high_sub_pixel_avg_variance16x16/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, const uint8_t *second_pred";
+ specialize qw/vp9_high_sub_pixel_avg_variance16x16/;
+
+ add_proto qw/unsigned int vp9_high_sub_pixel_variance8x16/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
+ specialize qw/vp9_high_sub_pixel_variance8x16/;
+
+ add_proto qw/unsigned int vp9_high_sub_pixel_avg_variance8x16/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, const uint8_t *second_pred";
+ specialize qw/vp9_high_sub_pixel_avg_variance8x16/;
+
+ add_proto qw/unsigned int vp9_high_sub_pixel_variance16x8/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
+ specialize qw/vp9_high_sub_pixel_variance16x8/;
+
+ add_proto qw/unsigned int vp9_high_sub_pixel_avg_variance16x8/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, const uint8_t *second_pred";
+ specialize qw/vp9_high_sub_pixel_avg_variance16x8/;
+
+ add_proto qw/unsigned int vp9_high_sub_pixel_variance8x8/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
+ specialize qw/vp9_high_sub_pixel_variance8x8/;
+
+ add_proto qw/unsigned int vp9_high_sub_pixel_avg_variance8x8/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, const uint8_t *second_pred";
+ specialize qw/vp9_high_sub_pixel_avg_variance8x8/;
+
+ add_proto qw/unsigned int vp9_high_sub_pixel_variance8x4/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
+ specialize qw/vp9_high_sub_pixel_variance8x4/;
+
+ add_proto qw/unsigned int vp9_high_sub_pixel_avg_variance8x4/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, const uint8_t *second_pred";
+ specialize qw/vp9_high_sub_pixel_avg_variance8x4/;
+
+ add_proto qw/unsigned int vp9_high_sub_pixel_variance4x8/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
+ specialize qw/vp9_high_sub_pixel_variance4x8/;
+
+ add_proto qw/unsigned int vp9_high_sub_pixel_avg_variance4x8/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, const uint8_t *second_pred";
+ specialize qw/vp9_high_sub_pixel_avg_variance4x8/;
+
+ add_proto qw/unsigned int vp9_high_sub_pixel_variance4x4/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
+ specialize qw/vp9_high_sub_pixel_variance4x4/;
+
+ add_proto qw/unsigned int vp9_high_sub_pixel_avg_variance4x4/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, const uint8_t *second_pred";
+ specialize qw/vp9_high_sub_pixel_avg_variance4x4/;
+
+ add_proto qw/unsigned int vp9_high_10_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 qw/vp9_high_10_sub_pixel_variance64x64/;
+
+ add_proto qw/unsigned int vp9_high_10_sub_pixel_avg_variance64x64/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, const uint8_t *second_pred";
+ specialize qw/vp9_high_10_sub_pixel_avg_variance64x64/;
+
+ add_proto qw/unsigned int vp9_high_10_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 qw/vp9_high_10_sub_pixel_variance32x64/;
+
+ add_proto qw/unsigned int vp9_high_10_sub_pixel_avg_variance32x64/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, const uint8_t *second_pred";
+ specialize qw/vp9_high_10_sub_pixel_avg_variance32x64/;
+
+ add_proto qw/unsigned int vp9_high_10_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 qw/vp9_high_10_sub_pixel_variance64x32/;
+
+ add_proto qw/unsigned int vp9_high_10_sub_pixel_avg_variance64x32/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, const uint8_t *second_pred";
+ specialize qw/vp9_high_10_sub_pixel_avg_variance64x32/;
+
+ add_proto qw/unsigned int vp9_high_10_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 qw/vp9_high_10_sub_pixel_variance32x16/;
+
+ add_proto qw/unsigned int vp9_high_10_sub_pixel_avg_variance32x16/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, const uint8_t *second_pred";
+ specialize qw/vp9_high_10_sub_pixel_avg_variance32x16/;
+
+ add_proto qw/unsigned int vp9_high_10_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 qw/vp9_high_10_sub_pixel_variance16x32/;
+
+ add_proto qw/unsigned int vp9_high_10_sub_pixel_avg_variance16x32/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, const uint8_t *second_pred";
+ specialize qw/vp9_high_10_sub_pixel_avg_variance16x32/;
+
+ add_proto qw/unsigned int vp9_high_10_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 qw/vp9_high_10_sub_pixel_variance32x32/;
+
+ add_proto qw/unsigned int vp9_high_10_sub_pixel_avg_variance32x32/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, const uint8_t *second_pred";
+ specialize qw/vp9_high_10_sub_pixel_avg_variance32x32/;
+
+ add_proto qw/unsigned int vp9_high_10_sub_pixel_variance16x16/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
+ specialize qw/vp9_high_10_sub_pixel_variance16x16/;
+
+ add_proto qw/unsigned int vp9_high_10_sub_pixel_avg_variance16x16/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, const uint8_t *second_pred";
+ specialize qw/vp9_high_10_sub_pixel_avg_variance16x16/;
+
+ add_proto qw/unsigned int vp9_high_10_sub_pixel_variance8x16/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
+ specialize qw/vp9_high_10_sub_pixel_variance8x16/;
+
+ add_proto qw/unsigned int vp9_high_10_sub_pixel_avg_variance8x16/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, const uint8_t *second_pred";
+ specialize qw/vp9_high_10_sub_pixel_avg_variance8x16/;
+
+ add_proto qw/unsigned int vp9_high_10_sub_pixel_variance16x8/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
+ specialize qw/vp9_high_10_sub_pixel_variance16x8/;
+
+ add_proto qw/unsigned int vp9_high_10_sub_pixel_avg_variance16x8/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, const uint8_t *second_pred";
+ specialize qw/vp9_high_10_sub_pixel_avg_variance16x8/;
+
+ add_proto qw/unsigned int vp9_high_10_sub_pixel_variance8x8/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
+ specialize qw/vp9_high_10_sub_pixel_variance8x8/;
+
+ add_proto qw/unsigned int vp9_high_10_sub_pixel_avg_variance8x8/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, const uint8_t *second_pred";
+ specialize qw/vp9_high_10_sub_pixel_avg_variance8x8/;
+
+ add_proto qw/unsigned int vp9_high_10_sub_pixel_variance8x4/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
+ specialize qw/vp9_high_10_sub_pixel_variance8x4/;
+
+ add_proto qw/unsigned int vp9_high_10_sub_pixel_avg_variance8x4/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, const uint8_t *second_pred";
+ specialize qw/vp9_high_10_sub_pixel_avg_variance8x4/;
+
+ add_proto qw/unsigned int vp9_high_10_sub_pixel_variance4x8/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
+ specialize qw/vp9_high_10_sub_pixel_variance4x8/;
+
+ add_proto qw/unsigned int vp9_high_10_sub_pixel_avg_variance4x8/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, const uint8_t *second_pred";
+ specialize qw/vp9_high_10_sub_pixel_avg_variance4x8/;
+
+ add_proto qw/unsigned int vp9_high_10_sub_pixel_variance4x4/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
+ specialize qw/vp9_high_10_sub_pixel_variance4x4/;
+
+ add_proto qw/unsigned int vp9_high_10_sub_pixel_avg_variance4x4/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, const uint8_t *second_pred";
+ specialize qw/vp9_high_10_sub_pixel_avg_variance4x4/;
+
+ add_proto qw/unsigned int vp9_high_12_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 qw/vp9_high_12_sub_pixel_variance64x64/;
+
+ add_proto qw/unsigned int vp9_high_12_sub_pixel_avg_variance64x64/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, const uint8_t *second_pred";
+ specialize qw/vp9_high_12_sub_pixel_avg_variance64x64/;
+
+ add_proto qw/unsigned int vp9_high_12_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 qw/vp9_high_12_sub_pixel_variance32x64/;
+
+ add_proto qw/unsigned int vp9_high_12_sub_pixel_avg_variance32x64/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, const uint8_t *second_pred";
+ specialize qw/vp9_high_12_sub_pixel_avg_variance32x64/;
+
+ add_proto qw/unsigned int vp9_high_12_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 qw/vp9_high_12_sub_pixel_variance64x32/;
+
+ add_proto qw/unsigned int vp9_high_12_sub_pixel_avg_variance64x32/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, const uint8_t *second_pred";
+ specialize qw/vp9_high_12_sub_pixel_avg_variance64x32/;
+
+ add_proto qw/unsigned int vp9_high_12_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 qw/vp9_high_12_sub_pixel_variance32x16/;
+
+ add_proto qw/unsigned int vp9_high_12_sub_pixel_avg_variance32x16/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, const uint8_t *second_pred";
+ specialize qw/vp9_high_12_sub_pixel_avg_variance32x16/;
+
+ add_proto qw/unsigned int vp9_high_12_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 qw/vp9_high_12_sub_pixel_variance16x32/;
+
+ add_proto qw/unsigned int vp9_high_12_sub_pixel_avg_variance16x32/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, const uint8_t *second_pred";
+ specialize qw/vp9_high_12_sub_pixel_avg_variance16x32/;
+
+ add_proto qw/unsigned int vp9_high_12_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 qw/vp9_high_12_sub_pixel_variance32x32/;
+
+ add_proto qw/unsigned int vp9_high_12_sub_pixel_avg_variance32x32/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, const uint8_t *second_pred";
+ specialize qw/vp9_high_12_sub_pixel_avg_variance32x32/;
+
+ add_proto qw/unsigned int vp9_high_12_sub_pixel_variance16x16/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
+ specialize qw/vp9_high_12_sub_pixel_variance16x16/;
+
+ add_proto qw/unsigned int vp9_high_12_sub_pixel_avg_variance16x16/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, const uint8_t *second_pred";
+ specialize qw/vp9_high_12_sub_pixel_avg_variance16x16/;
+
+ add_proto qw/unsigned int vp9_high_12_sub_pixel_variance8x16/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
+ specialize qw/vp9_high_12_sub_pixel_variance8x16/;
+
+ add_proto qw/unsigned int vp9_high_12_sub_pixel_avg_variance8x16/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, const uint8_t *second_pred";
+ specialize qw/vp9_high_12_sub_pixel_avg_variance8x16/;
+
+ add_proto qw/unsigned int vp9_high_12_sub_pixel_variance16x8/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
+ specialize qw/vp9_high_12_sub_pixel_variance16x8/;
+
+ add_proto qw/unsigned int vp9_high_12_sub_pixel_avg_variance16x8/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, const uint8_t *second_pred";
+ specialize qw/vp9_high_12_sub_pixel_avg_variance16x8/;
+
+ add_proto qw/unsigned int vp9_high_12_sub_pixel_variance8x8/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
+ specialize qw/vp9_high_12_sub_pixel_variance8x8/;
+
+ add_proto qw/unsigned int vp9_high_12_sub_pixel_avg_variance8x8/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, const uint8_t *second_pred";
+ specialize qw/vp9_high_12_sub_pixel_avg_variance8x8/;
+
+ add_proto qw/unsigned int vp9_high_12_sub_pixel_variance8x4/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
+ specialize qw/vp9_high_12_sub_pixel_variance8x4/;
+
+ add_proto qw/unsigned int vp9_high_12_sub_pixel_avg_variance8x4/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, const uint8_t *second_pred";
+ specialize qw/vp9_high_12_sub_pixel_avg_variance8x4/;
+
+ add_proto qw/unsigned int vp9_high_12_sub_pixel_variance4x8/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
+ specialize qw/vp9_high_12_sub_pixel_variance4x8/;
+
+ add_proto qw/unsigned int vp9_high_12_sub_pixel_avg_variance4x8/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, const uint8_t *second_pred";
+ specialize qw/vp9_high_12_sub_pixel_avg_variance4x8/;
+
+ add_proto qw/unsigned int vp9_high_12_sub_pixel_variance4x4/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse";
+ specialize qw/vp9_high_12_sub_pixel_variance4x4/;
+
+ add_proto qw/unsigned int vp9_high_12_sub_pixel_avg_variance4x4/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, const uint8_t *second_pred";
+ specialize qw/vp9_high_12_sub_pixel_avg_variance4x4/;
+
+ add_proto qw/unsigned int vp9_high_sad64x64/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride";
+ specialize qw/vp9_high_sad64x64/;
+
+ add_proto qw/unsigned int vp9_high_sad32x64/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride";
+ specialize qw/vp9_high_sad32x64/;
+
+ add_proto qw/unsigned int vp9_high_sad64x32/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride";
+ specialize qw/vp9_high_sad64x32/;
+
+ add_proto qw/unsigned int vp9_high_sad32x16/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride";
+ specialize qw/vp9_high_sad32x16/;
+
+ add_proto qw/unsigned int vp9_high_sad16x32/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride";
+ specialize qw/vp9_high_sad16x32/;
+
+ add_proto qw/unsigned int vp9_high_sad32x32/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride";
+ specialize qw/vp9_high_sad32x32/;
+
+ add_proto qw/unsigned int vp9_high_sad16x16/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride";
+ specialize qw/vp9_high_sad16x16/;
+
+ add_proto qw/unsigned int vp9_high_sad16x8/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride";
+ specialize qw/vp9_high_sad16x8/;
+
+ add_proto qw/unsigned int vp9_high_sad8x16/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride";
+ specialize qw/vp9_high_sad8x16/;
+
+ add_proto qw/unsigned int vp9_high_sad8x8/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride";
+ specialize qw/vp9_high_sad8x8/;
+
+ add_proto qw/unsigned int vp9_high_sad8x4/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride";
+ specialize qw/vp9_high_sad8x4/;
+
+ add_proto qw/unsigned int vp9_high_sad4x8/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride";
+ specialize qw/vp9_high_sad4x8/;
+
+ add_proto qw/unsigned int vp9_high_sad4x4/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride";
+ specialize qw/vp9_high_sad4x4/;
+
+ add_proto qw/unsigned int vp9_high_sad64x64_avg/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred";
+ specialize qw/vp9_high_sad64x64_avg/;
+
+ add_proto qw/unsigned int vp9_high_sad32x64_avg/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred";
+ specialize qw/vp9_high_sad32x64_avg/;
+
+ add_proto qw/unsigned int vp9_high_sad64x32_avg/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred";
+ specialize qw/vp9_high_sad64x32_avg/;
+
+ add_proto qw/unsigned int vp9_high_sad32x16_avg/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred";
+ specialize qw/vp9_high_sad32x16_avg/;
+
+ add_proto qw/unsigned int vp9_high_sad16x32_avg/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred";
+ specialize qw/vp9_high_sad16x32_avg/;
+
+ add_proto qw/unsigned int vp9_high_sad32x32_avg/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred";
+ specialize qw/vp9_high_sad32x32_avg/;
+
+ add_proto qw/unsigned int vp9_high_sad16x16_avg/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred";
+ specialize qw/vp9_high_sad16x16_avg/;
+
+ add_proto qw/unsigned int vp9_high_sad16x8_avg/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred";
+ specialize qw/vp9_high_sad16x8_avg/;
+
+ add_proto qw/unsigned int vp9_high_sad8x16_avg/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred";
+ specialize qw/vp9_high_sad8x16_avg/;
+
+ add_proto qw/unsigned int vp9_high_sad8x8_avg/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred";
+ specialize qw/vp9_high_sad8x8_avg/;
+
+ add_proto qw/unsigned int vp9_high_sad8x4_avg/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred";
+ specialize qw/vp9_high_sad8x4_avg/;
+
+ add_proto qw/unsigned int vp9_high_sad4x8_avg/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred";
+ specialize qw/vp9_high_sad4x8_avg/;
+
+ add_proto qw/unsigned int vp9_high_sad4x4_avg/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred";
+ specialize qw/vp9_high_sad4x4_avg/;
+
+ add_proto qw/void vp9_high_sad64x64x3/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sad_array";
+ specialize qw/vp9_high_sad64x64x3/;
+
+ add_proto qw/void vp9_high_sad32x32x3/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sad_array";
+ specialize qw/vp9_high_sad32x32x3/;
+
+ add_proto qw/void vp9_high_sad16x16x3/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sad_array";
+ specialize qw/vp9_high_sad16x16x3/;
+
+ add_proto qw/void vp9_high_sad16x8x3/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sad_array";
+ specialize qw/vp9_high_sad16x8x3/;
+
+ add_proto qw/void vp9_high_sad8x16x3/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sad_array";
+ specialize qw/vp9_high_sad8x16x3/;
+
+ add_proto qw/void vp9_high_sad8x8x3/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sad_array";
+ specialize qw/vp9_high_sad8x8x3/;
+
+ add_proto qw/void vp9_high_sad4x4x3/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sad_array";
+ specialize qw/vp9_high_sad4x4x3/;
+
+ add_proto qw/void vp9_high_sad64x64x8/, "const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, uint32_t *sad_array";
+ specialize qw/vp9_high_sad64x64x8/;
+
+ add_proto qw/void vp9_high_sad32x32x8/, "const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, uint32_t *sad_array";
+ specialize qw/vp9_high_sad32x32x8/;
+
+ add_proto qw/void vp9_high_sad16x16x8/, "const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, uint32_t *sad_array";
+ specialize qw/vp9_high_sad16x16x8/;
+
+ add_proto qw/void vp9_high_sad16x8x8/, "const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, uint32_t *sad_array";
+ specialize qw/vp9_high_sad16x8x8/;
+
+ add_proto qw/void vp9_high_sad8x16x8/, "const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, uint32_t *sad_array";
+ specialize qw/vp9_high_sad8x16x8/;
+
+ add_proto qw/void vp9_high_sad8x8x8/, "const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, uint32_t *sad_array";
+ specialize qw/vp9_high_sad8x8x8/;
+
+ add_proto qw/void vp9_high_sad8x4x8/, "const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, uint32_t *sad_array";
+ specialize qw/vp9_high_sad8x4x8/;
+
+ add_proto qw/void vp9_high_sad4x8x8/, "const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, uint32_t *sad_array";
+ specialize qw/vp9_high_sad4x8x8/;
+
+ add_proto qw/void vp9_high_sad4x4x8/, "const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, uint32_t *sad_array";
+ specialize qw/vp9_high_sad4x4x8/;
+
+ add_proto qw/void vp9_high_sad64x64x4d/, "const uint8_t *src_ptr, int src_stride, const uint8_t* const ref_ptr[], int ref_stride, unsigned int *sad_array";
+ specialize qw/vp9_high_sad64x64x4d/;
+
+ add_proto qw/void vp9_high_sad32x64x4d/, "const uint8_t *src_ptr, int src_stride, const uint8_t* const ref_ptr[], int ref_stride, unsigned int *sad_array";
+ specialize qw/vp9_high_sad32x64x4d/;
+
+ add_proto qw/void vp9_high_sad64x32x4d/, "const uint8_t *src_ptr, int src_stride, const uint8_t* const ref_ptr[], int ref_stride, unsigned int *sad_array";
+ specialize qw/vp9_high_sad64x32x4d/;
+
+ add_proto qw/void vp9_high_sad32x16x4d/, "const uint8_t *src_ptr, int src_stride, const uint8_t* const ref_ptr[], int ref_stride, unsigned int *sad_array";
+ specialize qw/vp9_high_sad32x16x4d/;
+
+ add_proto qw/void vp9_high_sad16x32x4d/, "const uint8_t *src_ptr, int src_stride, const uint8_t* const ref_ptr[], int ref_stride, unsigned int *sad_array";
+ specialize qw/vp9_high_sad16x32x4d/;
+
+ add_proto qw/void vp9_high_sad32x32x4d/, "const uint8_t *src_ptr, int src_stride, const uint8_t* const ref_ptr[], int ref_stride, unsigned int *sad_array";
+ specialize qw/vp9_high_sad32x32x4d/;
+
+ add_proto qw/void vp9_high_sad16x16x4d/, "const uint8_t *src_ptr, int src_stride, const uint8_t* const ref_ptr[], int ref_stride, unsigned int *sad_array";
+ specialize qw/vp9_high_sad16x16x4d/;
+
+ add_proto qw/void vp9_high_sad16x8x4d/, "const uint8_t *src_ptr, int src_stride, const uint8_t* const ref_ptr[], int ref_stride, unsigned int *sad_array";
+ specialize qw/vp9_high_sad16x8x4d/;
+
+ add_proto qw/void vp9_high_sad8x16x4d/, "const uint8_t *src_ptr, int src_stride, const uint8_t* const ref_ptr[], int ref_stride, unsigned int *sad_array";
+ specialize qw/vp9_high_sad8x16x4d/;
+
+ add_proto qw/void vp9_high_sad8x8x4d/, "const uint8_t *src_ptr, int src_stride, const uint8_t* const ref_ptr[], int ref_stride, unsigned int *sad_array";
+ specialize qw/vp9_high_sad8x8x4d/;
+
+ # TODO(jingning): need to convert these 4x8/8x4 functions into sse2 form
+ add_proto qw/void vp9_high_sad8x4x4d/, "const uint8_t *src_ptr, int src_stride, const uint8_t* const ref_ptr[], int ref_stride, unsigned int *sad_array";
+ specialize qw/vp9_high_sad8x4x4d/;
+
+ add_proto qw/void vp9_high_sad4x8x4d/, "const uint8_t *src_ptr, int src_stride, const uint8_t* const ref_ptr[], int ref_stride, unsigned int *sad_array";
+ specialize qw/vp9_high_sad4x8x4d/;
+
+ add_proto qw/void vp9_high_sad4x4x4d/, "const uint8_t *src_ptr, int src_stride, const uint8_t* const ref_ptr[], int ref_stride, unsigned int *sad_array";
+ specialize qw/vp9_high_sad4x4x4d/;
+
+ add_proto qw/unsigned int vp9_high_mse16x16/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int recon_stride, unsigned int *sse";
+ specialize qw/vp9_high_mse16x16/;
+
+ add_proto qw/unsigned int vp9_high_mse8x16/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int recon_stride, unsigned int *sse";
+ specialize qw/vp9_high_mse8x16/;
+
+ add_proto qw/unsigned int vp9_high_mse16x8/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int recon_stride, unsigned int *sse";
+ specialize qw/vp9_high_mse16x8/;
+
+ add_proto qw/unsigned int vp9_high_mse8x8/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int recon_stride, unsigned int *sse";
+ specialize qw/vp9_high_mse8x8/;
+
+ add_proto qw/unsigned int vp9_high_10_mse16x16/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int recon_stride, unsigned int *sse";
+ specialize qw/vp9_high_10_mse16x16/;
+
+ add_proto qw/unsigned int vp9_high_10_mse8x16/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int recon_stride, unsigned int *sse";
+ specialize qw/vp9_high_10_mse8x16/;
+
+ add_proto qw/unsigned int vp9_high_10_mse16x8/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int recon_stride, unsigned int *sse";
+ specialize qw/vp9_high_10_mse16x8/;
+
+ add_proto qw/unsigned int vp9_high_10_mse8x8/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int recon_stride, unsigned int *sse";
+ specialize qw/vp9_high_10_mse8x8/;
+
+ add_proto qw/unsigned int vp9_high_12_mse16x16/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int recon_stride, unsigned int *sse";
+ specialize qw/vp9_high_12_mse16x16/;
+
+ add_proto qw/unsigned int vp9_high_12_mse8x16/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int recon_stride, unsigned int *sse";
+ specialize qw/vp9_high_12_mse8x16/;
+
+ add_proto qw/unsigned int vp9_high_12_mse16x8/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int recon_stride, unsigned int *sse";
+ specialize qw/vp9_high_12_mse16x8/;
+
+ add_proto qw/unsigned int vp9_high_12_mse8x8/, "const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int recon_stride, unsigned int *sse";
+ specialize qw/vp9_high_12_mse8x8/;
+
+ # ENCODEMB INVOKE
+
+ add_proto qw/int64_t vp9_high_block_error/, "const tran_low_t *coeff, const tran_low_t *dqcoeff, intptr_t block_size, int64_t *ssz, int bd";
+ specialize qw/vp9_high_block_error/;
+
+ add_proto qw/void vp9_high_subtract_block/, "int rows, int cols, int16_t *diff_ptr, ptrdiff_t diff_stride, const uint8_t *src_ptr, ptrdiff_t src_stride, const uint8_t *pred_ptr, ptrdiff_t pred_stride, int bd";
+ specialize qw/vp9_high_subtract_block/;
+
+ add_proto qw/void vp9_high_quantize_fp/, "const tran_low_t *coeff_ptr, intptr_t n_coeffs, int skip_block, const int16_t *zbin_ptr, const int16_t *round_ptr, const int16_t *quant_ptr, const int16_t *quant_shift_ptr, tran_low_t *qcoeff_ptr, tran_low_t *dqcoeff_ptr, const int16_t *dequant_ptr, int zbin_oq_value, uint16_t *eob_ptr, const int16_t *scan, const int16_t *iscan";
+ specialize qw/vp9_high_quantize_fp/;
+
+ add_proto qw/void vp9_high_quantize_fp_32x32/, "const tran_low_t *coeff_ptr, intptr_t n_coeffs, int skip_block, const int16_t *zbin_ptr, const int16_t *round_ptr, const int16_t *quant_ptr, const int16_t *quant_shift_ptr, tran_low_t *qcoeff_ptr, tran_low_t *dqcoeff_ptr, const int16_t *dequant_ptr, int zbin_oq_value, uint16_t *eob_ptr, const int16_t *scan, const int16_t *iscan";
+ specialize qw/vp9_high_quantize_fp_32x32/;
+
+ add_proto qw/void vp9_high_quantize_b/, "const tran_low_t *coeff_ptr, intptr_t n_coeffs, int skip_block, const int16_t *zbin_ptr, const int16_t *round_ptr, const int16_t *quant_ptr, const int16_t *quant_shift_ptr, tran_low_t *qcoeff_ptr, tran_low_t *dqcoeff_ptr, const int16_t *dequant_ptr, int zbin_oq_value, uint16_t *eob_ptr, const int16_t *scan, const int16_t *iscan";
+ specialize qw/vp9_high_quantize_b/;
+
+ add_proto qw/void vp9_high_quantize_b_32x32/, "const tran_low_t *coeff_ptr, intptr_t n_coeffs, int skip_block, const int16_t *zbin_ptr, const int16_t *round_ptr, const int16_t *quant_ptr, const int16_t *quant_shift_ptr, tran_low_t *qcoeff_ptr, tran_low_t *dqcoeff_ptr, const int16_t *dequant_ptr, int zbin_oq_value, uint16_t *eob_ptr, const int16_t *scan, const int16_t *iscan";
+ specialize qw/vp9_high_quantize_b_32x32/;
+
+ #
+ # Structured Similarity (SSIM)
+ #
+ if (vpx_config("CONFIG_INTERNAL_STATS") eq "yes") {
+ add_proto qw/void vp9_high_ssim_parms_8x8/, "uint16_t *s, int sp, uint16_t *r, int rp, uint32_t *sum_s, uint32_t *sum_r, uint32_t *sum_sq_s, uint32_t *sum_sq_r, uint32_t *sum_sxr";
+ specialize qw/vp9_high_ssim_parms_8x8/;
+
+ add_proto qw/void vp9_high_ssim_parms_8x8_shift/, "uint16_t *s, int sp, uint16_t *r, int rp, uint32_t *sum_s, uint32_t *sum_r, uint32_t *sum_sq_s, uint32_t *sum_sq_r, uint32_t *sum_sxr, unsigned int bd, unsigned int shift";
+ specialize qw/vp9_high_ssim_parms_8x8_shift/;
+ }
+
+ # fdct functions
+ add_proto qw/void vp9_high_fht4x4/, "const int16_t *input, tran_low_t *output, int stride, int tx_type";
+ specialize qw/vp9_high_fht4x4/;
+
+ add_proto qw/void vp9_high_fht8x8/, "const int16_t *input, tran_low_t *output, int stride, int tx_type";
+ specialize qw/vp9_high_fht8x8/;
+
+ add_proto qw/void vp9_high_fht16x16/, "const int16_t *input, tran_low_t *output, int stride, int tx_type";
+ specialize qw/vp9_high_fht16x16/;
+
+ add_proto qw/void vp9_high_fwht4x4/, "const int16_t *input, tran_low_t *output, int stride";
+ specialize qw/vp9_high_fwht4x4/;
+
+ add_proto qw/void vp9_high_fdct4x4/, "const int16_t *input, tran_low_t *output, int stride";
+ specialize qw/vp9_high_fdct4x4/;
+
+ add_proto qw/void vp9_high_fdct8x8_1/, "const int16_t *input, tran_low_t *output, int stride";
+ specialize qw/vp9_high_fdct8x8_1/;
+
+ add_proto qw/void vp9_high_fdct8x8/, "const int16_t *input, tran_low_t *output, int stride";
+ specialize qw/vp9_high_fdct8x8/;
+
+ add_proto qw/void vp9_high_fdct16x16_1/, "const int16_t *input, tran_low_t *output, int stride";
+ specialize qw/vp9_high_fdct16x16_1/;
+
+ add_proto qw/void vp9_high_fdct16x16/, "const int16_t *input, tran_low_t *output, int stride";
+ specialize qw/vp9_high_fdct16x16/;
+
+ add_proto qw/void vp9_high_fdct32x32_1/, "const int16_t *input, tran_low_t *output, int stride";
+ specialize qw/vp9_high_fdct32x32_1/;
+
+ add_proto qw/void vp9_high_fdct32x32/, "const int16_t *input, tran_low_t *output, int stride";
+ specialize qw/vp9_high_fdct32x32/;
+
+ add_proto qw/void vp9_high_fdct32x32_rd/, "const int16_t *input, tran_low_t *output, int stride";
+ specialize qw/vp9_high_fdct32x32_rd/;
+
+ add_proto qw/void vp9_high_temporal_filter_apply/, "uint8_t *frame1, unsigned int stride, uint8_t *frame2, unsigned int block_width, unsigned int block_height, int strength, int filter_weight, unsigned int *accumulator, uint16_t *count";
+ specialize qw/vp9_high_temporal_filter_apply/;
+
}
+# End vp9_high encoder functions
+
+}
# end encoder functions
1;
« no previous file with comments | « source/libvpx/vp9/common/vp9_reconintra.c ('k') | source/libvpx/vp9/common/vp9_scale.h » ('j') | no next file with comments »

Powered by Google App Engine
This is Rietveld 408576698