mirror of
https://github.com/roytam1/palemoon26.git
synced 2026-05-26 05:44:31 +00:00
libvpx update
This commit is contained in:
@@ -288,6 +288,16 @@ CSRCS += \
|
||||
vp9_variance_sse2.c \
|
||||
variance_mmx.c \
|
||||
variance_sse2.c \
|
||||
vp9_loopfilter_intrin_avx2.c \
|
||||
vp9_subpixel_8t_intrin_avx2.c \
|
||||
vp9_dct_avx2.c \
|
||||
vp9_error_intrin_avx2.c \
|
||||
vp9_subpel_variance_impl_intrin_avx2.c \
|
||||
vp9_variance_avx2.c \
|
||||
sad4d_avx2.c \
|
||||
sad_avx2.c \
|
||||
variance_avx2.c \
|
||||
variance_impl_avx2.c \
|
||||
$(NULL)
|
||||
|
||||
VPX_ASM_ENC_OFFSETS_SRCS = \
|
||||
|
||||
+17
-2
@@ -398,14 +398,13 @@ def prepare_upstream(prefix, commit=None):
|
||||
configure = ['../../configure', '--target=%s' % target,
|
||||
'--disable-examples', '--disable-install-docs',
|
||||
'--enable-multi-res-encoding',
|
||||
'--size-limit=4000x3000'
|
||||
'--size-limit=8192x4608'
|
||||
]
|
||||
|
||||
if 'darwin9' in target:
|
||||
configure += ['--enable-pic']
|
||||
if 'linux' in target:
|
||||
configure += ['--enable-pic']
|
||||
configure += ['--disable-avx2']
|
||||
# x86inc.asm is not compatible with pic 32bit builds
|
||||
if target == 'x86-linux-gcc':
|
||||
configure += ['--disable-use-x86inc']
|
||||
@@ -593,6 +592,22 @@ def apply_patches():
|
||||
# Cherry pick https://chromium-review.googlesource.com/#/c/276889/
|
||||
# to fix crash on 32bit
|
||||
os.system("patch -p1 < vp9_filter_restore_aligment.patch")
|
||||
# Patch win32 vpx_once.
|
||||
os.system("patch -p3 < vpx_once.patch")
|
||||
# Bug 1224363 - Clamp seg_lvl also in abs-value mode.
|
||||
os.system("patch -p3 < clamp_abs_lvl_seg.patch")
|
||||
# Bug 1224361 - Clamp QIndex also in abs-value mode.
|
||||
os.system("patch -p3 < clamp-abs-QIndex.patch")
|
||||
# Bug 1233983 - Make libvpx build with clang-cl
|
||||
os.system("patch -p3 < clang-cl.patch")
|
||||
# Bug 1224371 - Cast uint8_t to uint32_t before shift
|
||||
os.system("patch -p3 < cast-char-to-uint-before-shift.patch")
|
||||
# Bug 1237848 - Check lookahead ctx
|
||||
os.system("patch -p3 < 1237848-check-lookahead-ctx.patch")
|
||||
# Bug 1263384 - Check input frame resolution
|
||||
os.system("patch -p3 < input_frame_validation.patch")
|
||||
# Bug 1315288 - Check input frame resolution for vp9
|
||||
os.system("patch -p3 < input_frame_validation_vp9.patch")
|
||||
|
||||
def update_readme(commit):
|
||||
with open('README_MOZILLA') as f:
|
||||
|
||||
@@ -24,6 +24,7 @@
|
||||
#include <unistd.h>
|
||||
#elif defined(_WIN32)
|
||||
#include <windows.h>
|
||||
#include <intrin.h>
|
||||
typedef void (WINAPI *PGNSI)(LPSYSTEM_INFO);
|
||||
#elif defined(__OS2__)
|
||||
#define INCL_DOS
|
||||
|
||||
@@ -11,6 +11,9 @@
|
||||
#define RTCD_C
|
||||
#include "./vp8_rtcd.h"
|
||||
#include "vpx_ports/vpx_once.h"
|
||||
#ifdef _MSC_VER
|
||||
#include <intrin.h>
|
||||
#endif
|
||||
|
||||
|
||||
void vp8_rtcd()
|
||||
|
||||
@@ -71,9 +71,8 @@ void vp8_mb_init_dequantizer(VP8D_COMP *pbi, MACROBLOCKD *xd)
|
||||
|
||||
/* Delta Value */
|
||||
else
|
||||
{
|
||||
QIndex = pc->base_qindex + xd->segment_feature_data[MB_LVL_ALT_Q][mbmi->segment_id];
|
||||
}
|
||||
|
||||
QIndex = (QIndex >= 0) ? ((QIndex <= MAXQ) ? QIndex : MAXQ) : 0; /* Clamp to valid range */
|
||||
}
|
||||
else
|
||||
|
||||
@@ -28,6 +28,9 @@
|
||||
#if CONFIG_ERROR_CONCEALMENT
|
||||
#include "error_concealment.h"
|
||||
#endif
|
||||
#ifdef _MSC_VER
|
||||
#include <intrin.h>
|
||||
#endif
|
||||
|
||||
#define CALLOC_ARRAY(p, n) CHECK_MEM_ERROR((p), vpx_calloc(sizeof(*(p)), (n)))
|
||||
#define CALLOC_ARRAY_ALIGNED(p, n, algn) do { \
|
||||
|
||||
@@ -34,6 +34,9 @@
|
||||
#include "bitstream.h"
|
||||
#endif
|
||||
#include "encodeframe.h"
|
||||
#ifdef _MSC_VER
|
||||
#include <intrin.h>
|
||||
#endif
|
||||
|
||||
extern void vp8_stuff_mb(VP8_COMP *cpi, MACROBLOCK *x, TOKENEXTRA **t) ;
|
||||
extern void vp8_calc_ref_frame_costs(int *ref_frame_cost,
|
||||
|
||||
@@ -14,6 +14,9 @@
|
||||
#include "vp8/common/extend.h"
|
||||
#include "bitstream.h"
|
||||
#include "encodeframe.h"
|
||||
#ifdef _MSC_VER
|
||||
#include <intrin.h>
|
||||
#endif
|
||||
|
||||
#if CONFIG_MULTITHREAD
|
||||
|
||||
|
||||
@@ -181,6 +181,7 @@ vp8_lookahead_pop(struct lookahead_ctx *ctx,
|
||||
{
|
||||
struct lookahead_entry* buf = NULL;
|
||||
|
||||
assert(ctx != NULL);
|
||||
if(ctx->sz && (drain || ctx->sz == ctx->max_sz - 1))
|
||||
{
|
||||
buf = pop(ctx, &ctx->read_idx);
|
||||
|
||||
@@ -922,7 +922,7 @@ static vpx_codec_err_t vp8e_encode(vpx_codec_alg_priv_t *ctx,
|
||||
res = image2yuvconfig(img, &sd);
|
||||
|
||||
if (sd.y_width != ctx->cfg.g_w || sd.y_height != ctx->cfg.g_h) {
|
||||
/* from vp8_encoder.h for g_w/g_h:
|
||||
/* from vpx_encoder.h for g_w/g_h:
|
||||
"Note that the frames passed as input to the encoder must have this resolution"
|
||||
*/
|
||||
ctx->base.err_detail = "Invalid input frame resolution";
|
||||
|
||||
@@ -499,7 +499,7 @@ vpx_codec_err_t vp9_parse_superframe_index(const uint8_t *data,
|
||||
uint32_t this_sz = 0;
|
||||
|
||||
for (j = 0; j < mag; ++j)
|
||||
this_sz |= (*x++) << (j * 8);
|
||||
this_sz |= (uint32_t)(*x++) << (j * 8);
|
||||
sizes[i] = this_sz;
|
||||
}
|
||||
*count = frames;
|
||||
|
||||
@@ -207,7 +207,7 @@ struct lookahead_entry *vp9_lookahead_pop(struct lookahead_ctx *ctx,
|
||||
int drain) {
|
||||
struct lookahead_entry *buf = NULL;
|
||||
|
||||
if (ctx->sz && (drain || ctx->sz == ctx->max_sz - MAX_PRE_FRAMES)) {
|
||||
if (ctx && ctx->sz && (drain || ctx->sz == ctx->max_sz - MAX_PRE_FRAMES)) {
|
||||
buf = pop(ctx, &ctx->read_idx);
|
||||
ctx->sz--;
|
||||
}
|
||||
|
||||
@@ -994,11 +994,19 @@ static vpx_codec_err_t encoder_encode(vpx_codec_alg_priv_t *ctx,
|
||||
if (img != NULL) {
|
||||
res = image2yuvconfig(img, &sd);
|
||||
|
||||
// Store the original flags in to the frame buffer. Will extract the
|
||||
// key frame flag when we actually encode this frame.
|
||||
if (vp9_receive_raw_frame(cpi, flags | ctx->next_frame_flags,
|
||||
&sd, dst_time_stamp, dst_end_time_stamp)) {
|
||||
res = update_error_state(ctx, &cpi->common.error);
|
||||
if (sd.y_width != ctx->cfg.g_w || sd.y_height != ctx->cfg.g_h) {
|
||||
/* from vpx_encoder.h for g_w/g_h:
|
||||
"Note that the frames passed as input to the encoder must have this resolution"
|
||||
*/
|
||||
ctx->base.err_detail = "Invalid input frame resolution";
|
||||
res = VPX_CODEC_INVALID_PARAM;
|
||||
} else {
|
||||
// Store the original flags in to the frame buffer. Will extract the
|
||||
// key frame flag when we actually encode this frame.
|
||||
if (vp9_receive_raw_frame(cpi, flags | ctx->next_frame_flags,
|
||||
&sd, dst_time_stamp, dst_end_time_stamp)) {
|
||||
res = update_error_state(ctx, &cpi->common.error);
|
||||
}
|
||||
}
|
||||
ctx->next_frame_flags = 0;
|
||||
}
|
||||
|
||||
@@ -38,7 +38,8 @@ unsigned int vp9_avg_8x8_sse2(const uint8_t *, int p);
|
||||
RTCD_EXTERN unsigned int (*vp9_avg_8x8)(const uint8_t *, int p);
|
||||
|
||||
int64_t vp9_block_error_c(const tran_low_t *coeff, const tran_low_t *dqcoeff, intptr_t block_size, int64_t *ssz);
|
||||
#define vp9_block_error vp9_block_error_c
|
||||
int64_t vp9_block_error_avx2(const tran_low_t *coeff, const tran_low_t *dqcoeff, intptr_t block_size, int64_t *ssz);
|
||||
RTCD_EXTERN int64_t (*vp9_block_error)(const tran_low_t *coeff, const tran_low_t *dqcoeff, intptr_t block_size, int64_t *ssz);
|
||||
|
||||
int64_t vp9_block_error_fp_c(const int16_t *coeff, const int16_t *dqcoeff, int block_size);
|
||||
int64_t vp9_block_error_fp_sse2(const int16_t *coeff, const int16_t *dqcoeff, int block_size);
|
||||
@@ -47,6 +48,7 @@ RTCD_EXTERN int64_t (*vp9_block_error_fp)(const int16_t *coeff, const int16_t *d
|
||||
void vp9_convolve8_c(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);
|
||||
void vp9_convolve8_sse2(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);
|
||||
void vp9_convolve8_ssse3(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);
|
||||
void vp9_convolve8_avx2(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);
|
||||
RTCD_EXTERN 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);
|
||||
|
||||
void vp9_convolve8_avg_c(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);
|
||||
@@ -67,11 +69,13 @@ RTCD_EXTERN void (*vp9_convolve8_avg_vert)(const uint8_t *src, ptrdiff_t src_str
|
||||
void vp9_convolve8_horiz_c(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);
|
||||
void vp9_convolve8_horiz_sse2(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);
|
||||
void vp9_convolve8_horiz_ssse3(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);
|
||||
void vp9_convolve8_horiz_avx2(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);
|
||||
RTCD_EXTERN 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);
|
||||
|
||||
void vp9_convolve8_vert_c(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);
|
||||
void vp9_convolve8_vert_sse2(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);
|
||||
void vp9_convolve8_vert_ssse3(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);
|
||||
void vp9_convolve8_vert_avx2(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);
|
||||
RTCD_EXTERN 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);
|
||||
|
||||
void vp9_convolve_avg_c(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);
|
||||
@@ -213,6 +217,7 @@ RTCD_EXTERN void (*vp9_fdct16x16_1)(const int16_t *input, tran_low_t *output, in
|
||||
|
||||
void vp9_fdct32x32_c(const int16_t *input, tran_low_t *output, int stride);
|
||||
void vp9_fdct32x32_sse2(const int16_t *input, tran_low_t *output, int stride);
|
||||
void vp9_fdct32x32_avx2(const int16_t *input, tran_low_t *output, int stride);
|
||||
RTCD_EXTERN void (*vp9_fdct32x32)(const int16_t *input, tran_low_t *output, int stride);
|
||||
|
||||
void vp9_fdct32x32_1_c(const int16_t *input, tran_low_t *output, int stride);
|
||||
@@ -221,6 +226,7 @@ RTCD_EXTERN void (*vp9_fdct32x32_1)(const int16_t *input, tran_low_t *output, in
|
||||
|
||||
void vp9_fdct32x32_rd_c(const int16_t *input, tran_low_t *output, int stride);
|
||||
void vp9_fdct32x32_rd_sse2(const int16_t *input, tran_low_t *output, int stride);
|
||||
void vp9_fdct32x32_rd_avx2(const int16_t *input, tran_low_t *output, int stride);
|
||||
RTCD_EXTERN void (*vp9_fdct32x32_rd)(const int16_t *input, tran_low_t *output, int stride);
|
||||
|
||||
void vp9_fdct4x4_c(const int16_t *input, tran_low_t *output, int stride);
|
||||
@@ -359,6 +365,7 @@ void vp9_iwht4x4_1_add_c(const tran_low_t *input, uint8_t *dest, int dest_stride
|
||||
|
||||
void vp9_lpf_horizontal_16_c(uint8_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh, int count);
|
||||
void vp9_lpf_horizontal_16_sse2(uint8_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh, int count);
|
||||
void vp9_lpf_horizontal_16_avx2(uint8_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh, int count);
|
||||
RTCD_EXTERN void (*vp9_lpf_horizontal_16)(uint8_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh, int count);
|
||||
|
||||
void vp9_lpf_horizontal_4_c(uint8_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh, int count);
|
||||
@@ -436,7 +443,8 @@ unsigned int vp9_sub_pixel_avg_variance32x16_c(const uint8_t *src_ptr, int sourc
|
||||
#define vp9_sub_pixel_avg_variance32x16 vp9_sub_pixel_avg_variance32x16_c
|
||||
|
||||
unsigned int vp9_sub_pixel_avg_variance32x32_c(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);
|
||||
#define vp9_sub_pixel_avg_variance32x32 vp9_sub_pixel_avg_variance32x32_c
|
||||
unsigned int vp9_sub_pixel_avg_variance32x32_avx2(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);
|
||||
RTCD_EXTERN unsigned int (*vp9_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);
|
||||
|
||||
unsigned int vp9_sub_pixel_avg_variance32x64_c(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);
|
||||
#define vp9_sub_pixel_avg_variance32x64 vp9_sub_pixel_avg_variance32x64_c
|
||||
@@ -451,7 +459,8 @@ unsigned int vp9_sub_pixel_avg_variance64x32_c(const uint8_t *src_ptr, int sourc
|
||||
#define vp9_sub_pixel_avg_variance64x32 vp9_sub_pixel_avg_variance64x32_c
|
||||
|
||||
unsigned int vp9_sub_pixel_avg_variance64x64_c(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);
|
||||
#define vp9_sub_pixel_avg_variance64x64 vp9_sub_pixel_avg_variance64x64_c
|
||||
unsigned int vp9_sub_pixel_avg_variance64x64_avx2(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);
|
||||
RTCD_EXTERN unsigned int (*vp9_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);
|
||||
|
||||
unsigned int vp9_sub_pixel_avg_variance8x16_c(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);
|
||||
#define vp9_sub_pixel_avg_variance8x16 vp9_sub_pixel_avg_variance8x16_c
|
||||
@@ -475,7 +484,8 @@ unsigned int vp9_sub_pixel_variance32x16_c(const uint8_t *src_ptr, int source_st
|
||||
#define vp9_sub_pixel_variance32x16 vp9_sub_pixel_variance32x16_c
|
||||
|
||||
unsigned int vp9_sub_pixel_variance32x32_c(const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
#define vp9_sub_pixel_variance32x32 vp9_sub_pixel_variance32x32_c
|
||||
unsigned int vp9_sub_pixel_variance32x32_avx2(const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
RTCD_EXTERN unsigned int (*vp9_sub_pixel_variance32x32)(const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
|
||||
unsigned int vp9_sub_pixel_variance32x64_c(const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
#define vp9_sub_pixel_variance32x64 vp9_sub_pixel_variance32x64_c
|
||||
@@ -490,7 +500,8 @@ unsigned int vp9_sub_pixel_variance64x32_c(const uint8_t *src_ptr, int source_st
|
||||
#define vp9_sub_pixel_variance64x32 vp9_sub_pixel_variance64x32_c
|
||||
|
||||
unsigned int vp9_sub_pixel_variance64x64_c(const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
#define vp9_sub_pixel_variance64x64 vp9_sub_pixel_variance64x64_c
|
||||
unsigned int vp9_sub_pixel_variance64x64_avx2(const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
RTCD_EXTERN unsigned int (*vp9_sub_pixel_variance64x64)(const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
|
||||
unsigned int vp9_sub_pixel_variance8x16_c(const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
#define vp9_sub_pixel_variance8x16 vp9_sub_pixel_variance8x16_c
|
||||
@@ -550,11 +561,14 @@ static void setup_rtcd_internal(void)
|
||||
if (flags & HAS_SSE2) vp9_avg_4x4 = vp9_avg_4x4_sse2;
|
||||
vp9_avg_8x8 = vp9_avg_8x8_c;
|
||||
if (flags & HAS_SSE2) vp9_avg_8x8 = vp9_avg_8x8_sse2;
|
||||
vp9_block_error = vp9_block_error_c;
|
||||
if (flags & HAS_AVX2) vp9_block_error = vp9_block_error_avx2;
|
||||
vp9_block_error_fp = vp9_block_error_fp_c;
|
||||
if (flags & HAS_SSE2) vp9_block_error_fp = vp9_block_error_fp_sse2;
|
||||
vp9_convolve8 = vp9_convolve8_c;
|
||||
if (flags & HAS_SSE2) vp9_convolve8 = vp9_convolve8_sse2;
|
||||
if (flags & HAS_SSSE3) vp9_convolve8 = vp9_convolve8_ssse3;
|
||||
if (flags & HAS_AVX2) vp9_convolve8 = vp9_convolve8_avx2;
|
||||
vp9_convolve8_avg = vp9_convolve8_avg_c;
|
||||
if (flags & HAS_SSE2) vp9_convolve8_avg = vp9_convolve8_avg_sse2;
|
||||
if (flags & HAS_SSSE3) vp9_convolve8_avg = vp9_convolve8_avg_ssse3;
|
||||
@@ -567,19 +581,23 @@ static void setup_rtcd_internal(void)
|
||||
vp9_convolve8_horiz = vp9_convolve8_horiz_c;
|
||||
if (flags & HAS_SSE2) vp9_convolve8_horiz = vp9_convolve8_horiz_sse2;
|
||||
if (flags & HAS_SSSE3) vp9_convolve8_horiz = vp9_convolve8_horiz_ssse3;
|
||||
if (flags & HAS_AVX2) vp9_convolve8_horiz = vp9_convolve8_horiz_avx2;
|
||||
vp9_convolve8_vert = vp9_convolve8_vert_c;
|
||||
if (flags & HAS_SSE2) vp9_convolve8_vert = vp9_convolve8_vert_sse2;
|
||||
if (flags & HAS_SSSE3) vp9_convolve8_vert = vp9_convolve8_vert_ssse3;
|
||||
if (flags & HAS_AVX2) vp9_convolve8_vert = vp9_convolve8_vert_avx2;
|
||||
vp9_fdct16x16 = vp9_fdct16x16_c;
|
||||
if (flags & HAS_SSE2) vp9_fdct16x16 = vp9_fdct16x16_sse2;
|
||||
vp9_fdct16x16_1 = vp9_fdct16x16_1_c;
|
||||
if (flags & HAS_SSE2) vp9_fdct16x16_1 = vp9_fdct16x16_1_sse2;
|
||||
vp9_fdct32x32 = vp9_fdct32x32_c;
|
||||
if (flags & HAS_SSE2) vp9_fdct32x32 = vp9_fdct32x32_sse2;
|
||||
if (flags & HAS_AVX2) vp9_fdct32x32 = vp9_fdct32x32_avx2;
|
||||
vp9_fdct32x32_1 = vp9_fdct32x32_1_c;
|
||||
if (flags & HAS_SSE2) vp9_fdct32x32_1 = vp9_fdct32x32_1_sse2;
|
||||
vp9_fdct32x32_rd = vp9_fdct32x32_rd_c;
|
||||
if (flags & HAS_SSE2) vp9_fdct32x32_rd = vp9_fdct32x32_rd_sse2;
|
||||
if (flags & HAS_AVX2) vp9_fdct32x32_rd = vp9_fdct32x32_rd_avx2;
|
||||
vp9_fdct4x4 = vp9_fdct4x4_c;
|
||||
if (flags & HAS_SSE2) vp9_fdct4x4 = vp9_fdct4x4_sse2;
|
||||
vp9_fdct4x4_1 = vp9_fdct4x4_1_c;
|
||||
@@ -638,6 +656,7 @@ static void setup_rtcd_internal(void)
|
||||
if (flags & HAS_SSE2) vp9_int_pro_row = vp9_int_pro_row_sse2;
|
||||
vp9_lpf_horizontal_16 = vp9_lpf_horizontal_16_c;
|
||||
if (flags & HAS_SSE2) vp9_lpf_horizontal_16 = vp9_lpf_horizontal_16_sse2;
|
||||
if (flags & HAS_AVX2) vp9_lpf_horizontal_16 = vp9_lpf_horizontal_16_avx2;
|
||||
vp9_lpf_horizontal_4 = vp9_lpf_horizontal_4_c;
|
||||
if (flags & HAS_MMX) vp9_lpf_horizontal_4 = vp9_lpf_horizontal_4_mmx;
|
||||
vp9_lpf_horizontal_4_dual = vp9_lpf_horizontal_4_dual_c;
|
||||
@@ -666,6 +685,14 @@ static void setup_rtcd_internal(void)
|
||||
if (flags & HAS_SSE2) vp9_quantize_fp = vp9_quantize_fp_sse2;
|
||||
vp9_satd = vp9_satd_c;
|
||||
if (flags & HAS_SSE2) vp9_satd = vp9_satd_sse2;
|
||||
vp9_sub_pixel_avg_variance32x32 = vp9_sub_pixel_avg_variance32x32_c;
|
||||
if (flags & HAS_AVX2) vp9_sub_pixel_avg_variance32x32 = vp9_sub_pixel_avg_variance32x32_avx2;
|
||||
vp9_sub_pixel_avg_variance64x64 = vp9_sub_pixel_avg_variance64x64_c;
|
||||
if (flags & HAS_AVX2) vp9_sub_pixel_avg_variance64x64 = vp9_sub_pixel_avg_variance64x64_avx2;
|
||||
vp9_sub_pixel_variance32x32 = vp9_sub_pixel_variance32x32_c;
|
||||
if (flags & HAS_AVX2) vp9_sub_pixel_variance32x32 = vp9_sub_pixel_variance32x32_avx2;
|
||||
vp9_sub_pixel_variance64x64 = vp9_sub_pixel_variance64x64_c;
|
||||
if (flags & HAS_AVX2) vp9_sub_pixel_variance64x64 = vp9_sub_pixel_variance64x64_avx2;
|
||||
vp9_temporal_filter_apply = vp9_temporal_filter_apply_c;
|
||||
if (flags & HAS_SSE2) vp9_temporal_filter_apply = vp9_temporal_filter_apply_sse2;
|
||||
vp9_vector_var = vp9_vector_var_c;
|
||||
|
||||
@@ -39,6 +39,7 @@ RTCD_EXTERN unsigned int (*vp9_avg_8x8)(const uint8_t *, int p);
|
||||
|
||||
int64_t vp9_block_error_c(const tran_low_t *coeff, const tran_low_t *dqcoeff, intptr_t block_size, int64_t *ssz);
|
||||
int64_t vp9_block_error_sse2(const tran_low_t *coeff, const tran_low_t *dqcoeff, intptr_t block_size, int64_t *ssz);
|
||||
int64_t vp9_block_error_avx2(const tran_low_t *coeff, const tran_low_t *dqcoeff, intptr_t block_size, int64_t *ssz);
|
||||
RTCD_EXTERN int64_t (*vp9_block_error)(const tran_low_t *coeff, const tran_low_t *dqcoeff, intptr_t block_size, int64_t *ssz);
|
||||
|
||||
int64_t vp9_block_error_fp_c(const int16_t *coeff, const int16_t *dqcoeff, int block_size);
|
||||
@@ -48,6 +49,7 @@ RTCD_EXTERN int64_t (*vp9_block_error_fp)(const int16_t *coeff, const int16_t *d
|
||||
void vp9_convolve8_c(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);
|
||||
void vp9_convolve8_sse2(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);
|
||||
void vp9_convolve8_ssse3(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);
|
||||
void vp9_convolve8_avx2(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);
|
||||
RTCD_EXTERN 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);
|
||||
|
||||
void vp9_convolve8_avg_c(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);
|
||||
@@ -68,11 +70,13 @@ RTCD_EXTERN void (*vp9_convolve8_avg_vert)(const uint8_t *src, ptrdiff_t src_str
|
||||
void vp9_convolve8_horiz_c(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);
|
||||
void vp9_convolve8_horiz_sse2(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);
|
||||
void vp9_convolve8_horiz_ssse3(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);
|
||||
void vp9_convolve8_horiz_avx2(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);
|
||||
RTCD_EXTERN 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);
|
||||
|
||||
void vp9_convolve8_vert_c(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);
|
||||
void vp9_convolve8_vert_sse2(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);
|
||||
void vp9_convolve8_vert_ssse3(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);
|
||||
void vp9_convolve8_vert_avx2(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);
|
||||
RTCD_EXTERN 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);
|
||||
|
||||
void vp9_convolve_avg_c(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);
|
||||
@@ -247,6 +251,7 @@ RTCD_EXTERN void (*vp9_fdct16x16_1)(const int16_t *input, tran_low_t *output, in
|
||||
|
||||
void vp9_fdct32x32_c(const int16_t *input, tran_low_t *output, int stride);
|
||||
void vp9_fdct32x32_sse2(const int16_t *input, tran_low_t *output, int stride);
|
||||
void vp9_fdct32x32_avx2(const int16_t *input, tran_low_t *output, int stride);
|
||||
RTCD_EXTERN void (*vp9_fdct32x32)(const int16_t *input, tran_low_t *output, int stride);
|
||||
|
||||
void vp9_fdct32x32_1_c(const int16_t *input, tran_low_t *output, int stride);
|
||||
@@ -255,6 +260,7 @@ RTCD_EXTERN void (*vp9_fdct32x32_1)(const int16_t *input, tran_low_t *output, in
|
||||
|
||||
void vp9_fdct32x32_rd_c(const int16_t *input, tran_low_t *output, int stride);
|
||||
void vp9_fdct32x32_rd_sse2(const int16_t *input, tran_low_t *output, int stride);
|
||||
void vp9_fdct32x32_rd_avx2(const int16_t *input, tran_low_t *output, int stride);
|
||||
RTCD_EXTERN void (*vp9_fdct32x32_rd)(const int16_t *input, tran_low_t *output, int stride);
|
||||
|
||||
void vp9_fdct4x4_c(const int16_t *input, tran_low_t *output, int stride);
|
||||
@@ -398,6 +404,7 @@ void vp9_iwht4x4_1_add_c(const tran_low_t *input, uint8_t *dest, int dest_stride
|
||||
|
||||
void vp9_lpf_horizontal_16_c(uint8_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh, int count);
|
||||
void vp9_lpf_horizontal_16_sse2(uint8_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh, int count);
|
||||
void vp9_lpf_horizontal_16_avx2(uint8_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh, int count);
|
||||
RTCD_EXTERN void (*vp9_lpf_horizontal_16)(uint8_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh, int count);
|
||||
|
||||
void vp9_lpf_horizontal_4_c(uint8_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh, int count);
|
||||
@@ -485,6 +492,7 @@ RTCD_EXTERN unsigned int (*vp9_sub_pixel_avg_variance32x16)(const uint8_t *src_p
|
||||
unsigned int vp9_sub_pixel_avg_variance32x32_c(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);
|
||||
unsigned int vp9_sub_pixel_avg_variance32x32_sse2(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);
|
||||
unsigned int vp9_sub_pixel_avg_variance32x32_ssse3(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);
|
||||
unsigned int vp9_sub_pixel_avg_variance32x32_avx2(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);
|
||||
RTCD_EXTERN unsigned int (*vp9_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);
|
||||
|
||||
unsigned int vp9_sub_pixel_avg_variance32x64_c(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);
|
||||
@@ -510,6 +518,7 @@ RTCD_EXTERN unsigned int (*vp9_sub_pixel_avg_variance64x32)(const uint8_t *src_p
|
||||
unsigned int vp9_sub_pixel_avg_variance64x64_c(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);
|
||||
unsigned int vp9_sub_pixel_avg_variance64x64_sse2(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);
|
||||
unsigned int vp9_sub_pixel_avg_variance64x64_ssse3(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);
|
||||
unsigned int vp9_sub_pixel_avg_variance64x64_avx2(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);
|
||||
RTCD_EXTERN unsigned int (*vp9_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);
|
||||
|
||||
unsigned int vp9_sub_pixel_avg_variance8x16_c(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);
|
||||
@@ -550,6 +559,7 @@ RTCD_EXTERN unsigned int (*vp9_sub_pixel_variance32x16)(const uint8_t *src_ptr,
|
||||
unsigned int vp9_sub_pixel_variance32x32_c(const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
unsigned int vp9_sub_pixel_variance32x32_sse2(const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
unsigned int vp9_sub_pixel_variance32x32_ssse3(const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
unsigned int vp9_sub_pixel_variance32x32_avx2(const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
RTCD_EXTERN unsigned int (*vp9_sub_pixel_variance32x32)(const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
|
||||
unsigned int vp9_sub_pixel_variance32x64_c(const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
@@ -575,6 +585,7 @@ RTCD_EXTERN unsigned int (*vp9_sub_pixel_variance64x32)(const uint8_t *src_ptr,
|
||||
unsigned int vp9_sub_pixel_variance64x64_c(const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
unsigned int vp9_sub_pixel_variance64x64_sse2(const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
unsigned int vp9_sub_pixel_variance64x64_ssse3(const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
unsigned int vp9_sub_pixel_variance64x64_avx2(const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
RTCD_EXTERN unsigned int (*vp9_sub_pixel_variance64x64)(const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
|
||||
unsigned int vp9_sub_pixel_variance8x16_c(const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
@@ -651,11 +662,13 @@ static void setup_rtcd_internal(void)
|
||||
if (flags & HAS_SSE2) vp9_avg_8x8 = vp9_avg_8x8_sse2;
|
||||
vp9_block_error = vp9_block_error_c;
|
||||
if (flags & HAS_SSE2) vp9_block_error = vp9_block_error_sse2;
|
||||
if (flags & HAS_AVX2) vp9_block_error = vp9_block_error_avx2;
|
||||
vp9_block_error_fp = vp9_block_error_fp_c;
|
||||
if (flags & HAS_SSE2) vp9_block_error_fp = vp9_block_error_fp_sse2;
|
||||
vp9_convolve8 = vp9_convolve8_c;
|
||||
if (flags & HAS_SSE2) vp9_convolve8 = vp9_convolve8_sse2;
|
||||
if (flags & HAS_SSSE3) vp9_convolve8 = vp9_convolve8_ssse3;
|
||||
if (flags & HAS_AVX2) vp9_convolve8 = vp9_convolve8_avx2;
|
||||
vp9_convolve8_avg = vp9_convolve8_avg_c;
|
||||
if (flags & HAS_SSE2) vp9_convolve8_avg = vp9_convolve8_avg_sse2;
|
||||
if (flags & HAS_SSSE3) vp9_convolve8_avg = vp9_convolve8_avg_ssse3;
|
||||
@@ -668,9 +681,11 @@ static void setup_rtcd_internal(void)
|
||||
vp9_convolve8_horiz = vp9_convolve8_horiz_c;
|
||||
if (flags & HAS_SSE2) vp9_convolve8_horiz = vp9_convolve8_horiz_sse2;
|
||||
if (flags & HAS_SSSE3) vp9_convolve8_horiz = vp9_convolve8_horiz_ssse3;
|
||||
if (flags & HAS_AVX2) vp9_convolve8_horiz = vp9_convolve8_horiz_avx2;
|
||||
vp9_convolve8_vert = vp9_convolve8_vert_c;
|
||||
if (flags & HAS_SSE2) vp9_convolve8_vert = vp9_convolve8_vert_sse2;
|
||||
if (flags & HAS_SSSE3) vp9_convolve8_vert = vp9_convolve8_vert_ssse3;
|
||||
if (flags & HAS_AVX2) vp9_convolve8_vert = vp9_convolve8_vert_avx2;
|
||||
vp9_convolve_avg = vp9_convolve_avg_c;
|
||||
if (flags & HAS_SSE2) vp9_convolve_avg = vp9_convolve_avg_sse2;
|
||||
vp9_convolve_copy = vp9_convolve_copy_c;
|
||||
@@ -743,10 +758,12 @@ static void setup_rtcd_internal(void)
|
||||
if (flags & HAS_SSE2) vp9_fdct16x16_1 = vp9_fdct16x16_1_sse2;
|
||||
vp9_fdct32x32 = vp9_fdct32x32_c;
|
||||
if (flags & HAS_SSE2) vp9_fdct32x32 = vp9_fdct32x32_sse2;
|
||||
if (flags & HAS_AVX2) vp9_fdct32x32 = vp9_fdct32x32_avx2;
|
||||
vp9_fdct32x32_1 = vp9_fdct32x32_1_c;
|
||||
if (flags & HAS_SSE2) vp9_fdct32x32_1 = vp9_fdct32x32_1_sse2;
|
||||
vp9_fdct32x32_rd = vp9_fdct32x32_rd_c;
|
||||
if (flags & HAS_SSE2) vp9_fdct32x32_rd = vp9_fdct32x32_rd_sse2;
|
||||
if (flags & HAS_AVX2) vp9_fdct32x32_rd = vp9_fdct32x32_rd_avx2;
|
||||
vp9_fdct4x4 = vp9_fdct4x4_c;
|
||||
if (flags & HAS_SSE2) vp9_fdct4x4 = vp9_fdct4x4_sse2;
|
||||
vp9_fdct4x4_1 = vp9_fdct4x4_1_c;
|
||||
@@ -815,6 +832,7 @@ static void setup_rtcd_internal(void)
|
||||
if (flags & HAS_SSE2) vp9_int_pro_row = vp9_int_pro_row_sse2;
|
||||
vp9_lpf_horizontal_16 = vp9_lpf_horizontal_16_c;
|
||||
if (flags & HAS_SSE2) vp9_lpf_horizontal_16 = vp9_lpf_horizontal_16_sse2;
|
||||
if (flags & HAS_AVX2) vp9_lpf_horizontal_16 = vp9_lpf_horizontal_16_avx2;
|
||||
vp9_lpf_horizontal_4 = vp9_lpf_horizontal_4_c;
|
||||
if (flags & HAS_MMX) vp9_lpf_horizontal_4 = vp9_lpf_horizontal_4_mmx;
|
||||
vp9_lpf_horizontal_4_dual = vp9_lpf_horizontal_4_dual_c;
|
||||
@@ -858,6 +876,7 @@ static void setup_rtcd_internal(void)
|
||||
vp9_sub_pixel_avg_variance32x32 = vp9_sub_pixel_avg_variance32x32_c;
|
||||
if (flags & HAS_SSE2) vp9_sub_pixel_avg_variance32x32 = vp9_sub_pixel_avg_variance32x32_sse2;
|
||||
if (flags & HAS_SSSE3) vp9_sub_pixel_avg_variance32x32 = vp9_sub_pixel_avg_variance32x32_ssse3;
|
||||
if (flags & HAS_AVX2) vp9_sub_pixel_avg_variance32x32 = vp9_sub_pixel_avg_variance32x32_avx2;
|
||||
vp9_sub_pixel_avg_variance32x64 = vp9_sub_pixel_avg_variance32x64_c;
|
||||
if (flags & HAS_SSE2) vp9_sub_pixel_avg_variance32x64 = vp9_sub_pixel_avg_variance32x64_sse2;
|
||||
if (flags & HAS_SSSE3) vp9_sub_pixel_avg_variance32x64 = vp9_sub_pixel_avg_variance32x64_ssse3;
|
||||
@@ -873,6 +892,7 @@ static void setup_rtcd_internal(void)
|
||||
vp9_sub_pixel_avg_variance64x64 = vp9_sub_pixel_avg_variance64x64_c;
|
||||
if (flags & HAS_SSE2) vp9_sub_pixel_avg_variance64x64 = vp9_sub_pixel_avg_variance64x64_sse2;
|
||||
if (flags & HAS_SSSE3) vp9_sub_pixel_avg_variance64x64 = vp9_sub_pixel_avg_variance64x64_ssse3;
|
||||
if (flags & HAS_AVX2) vp9_sub_pixel_avg_variance64x64 = vp9_sub_pixel_avg_variance64x64_avx2;
|
||||
vp9_sub_pixel_avg_variance8x16 = vp9_sub_pixel_avg_variance8x16_c;
|
||||
if (flags & HAS_SSE2) vp9_sub_pixel_avg_variance8x16 = vp9_sub_pixel_avg_variance8x16_sse2;
|
||||
if (flags & HAS_SSSE3) vp9_sub_pixel_avg_variance8x16 = vp9_sub_pixel_avg_variance8x16_ssse3;
|
||||
@@ -897,6 +917,7 @@ static void setup_rtcd_internal(void)
|
||||
vp9_sub_pixel_variance32x32 = vp9_sub_pixel_variance32x32_c;
|
||||
if (flags & HAS_SSE2) vp9_sub_pixel_variance32x32 = vp9_sub_pixel_variance32x32_sse2;
|
||||
if (flags & HAS_SSSE3) vp9_sub_pixel_variance32x32 = vp9_sub_pixel_variance32x32_ssse3;
|
||||
if (flags & HAS_AVX2) vp9_sub_pixel_variance32x32 = vp9_sub_pixel_variance32x32_avx2;
|
||||
vp9_sub_pixel_variance32x64 = vp9_sub_pixel_variance32x64_c;
|
||||
if (flags & HAS_SSE2) vp9_sub_pixel_variance32x64 = vp9_sub_pixel_variance32x64_sse2;
|
||||
if (flags & HAS_SSSE3) vp9_sub_pixel_variance32x64 = vp9_sub_pixel_variance32x64_ssse3;
|
||||
@@ -912,6 +933,7 @@ static void setup_rtcd_internal(void)
|
||||
vp9_sub_pixel_variance64x64 = vp9_sub_pixel_variance64x64_c;
|
||||
if (flags & HAS_SSE2) vp9_sub_pixel_variance64x64 = vp9_sub_pixel_variance64x64_sse2;
|
||||
if (flags & HAS_SSSE3) vp9_sub_pixel_variance64x64 = vp9_sub_pixel_variance64x64_ssse3;
|
||||
if (flags & HAS_AVX2) vp9_sub_pixel_variance64x64 = vp9_sub_pixel_variance64x64_avx2;
|
||||
vp9_sub_pixel_variance8x16 = vp9_sub_pixel_variance8x16_c;
|
||||
if (flags & HAS_SSE2) vp9_sub_pixel_variance8x16 = vp9_sub_pixel_variance8x16_sse2;
|
||||
if (flags & HAS_SSSE3) vp9_sub_pixel_variance8x16 = vp9_sub_pixel_variance8x16_ssse3;
|
||||
|
||||
@@ -39,7 +39,8 @@ unsigned int vp9_avg_8x8_sse2(const uint8_t *, int p);
|
||||
|
||||
int64_t vp9_block_error_c(const tran_low_t *coeff, const tran_low_t *dqcoeff, intptr_t block_size, int64_t *ssz);
|
||||
int64_t vp9_block_error_sse2(const tran_low_t *coeff, const tran_low_t *dqcoeff, intptr_t block_size, int64_t *ssz);
|
||||
#define vp9_block_error vp9_block_error_sse2
|
||||
int64_t vp9_block_error_avx2(const tran_low_t *coeff, const tran_low_t *dqcoeff, intptr_t block_size, int64_t *ssz);
|
||||
RTCD_EXTERN int64_t (*vp9_block_error)(const tran_low_t *coeff, const tran_low_t *dqcoeff, intptr_t block_size, int64_t *ssz);
|
||||
|
||||
int64_t vp9_block_error_fp_c(const int16_t *coeff, const int16_t *dqcoeff, int block_size);
|
||||
int64_t vp9_block_error_fp_sse2(const int16_t *coeff, const int16_t *dqcoeff, int block_size);
|
||||
@@ -48,6 +49,7 @@ int64_t vp9_block_error_fp_sse2(const int16_t *coeff, const int16_t *dqcoeff, in
|
||||
void vp9_convolve8_c(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);
|
||||
void vp9_convolve8_sse2(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);
|
||||
void vp9_convolve8_ssse3(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);
|
||||
void vp9_convolve8_avx2(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);
|
||||
RTCD_EXTERN 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);
|
||||
|
||||
void vp9_convolve8_avg_c(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);
|
||||
@@ -68,11 +70,13 @@ RTCD_EXTERN void (*vp9_convolve8_avg_vert)(const uint8_t *src, ptrdiff_t src_str
|
||||
void vp9_convolve8_horiz_c(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);
|
||||
void vp9_convolve8_horiz_sse2(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);
|
||||
void vp9_convolve8_horiz_ssse3(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);
|
||||
void vp9_convolve8_horiz_avx2(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);
|
||||
RTCD_EXTERN 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);
|
||||
|
||||
void vp9_convolve8_vert_c(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);
|
||||
void vp9_convolve8_vert_sse2(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);
|
||||
void vp9_convolve8_vert_ssse3(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);
|
||||
void vp9_convolve8_vert_avx2(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);
|
||||
RTCD_EXTERN 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);
|
||||
|
||||
void vp9_convolve_avg_c(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);
|
||||
@@ -247,7 +251,8 @@ void vp9_fdct16x16_1_sse2(const int16_t *input, tran_low_t *output, int stride);
|
||||
|
||||
void vp9_fdct32x32_c(const int16_t *input, tran_low_t *output, int stride);
|
||||
void vp9_fdct32x32_sse2(const int16_t *input, tran_low_t *output, int stride);
|
||||
#define vp9_fdct32x32 vp9_fdct32x32_sse2
|
||||
void vp9_fdct32x32_avx2(const int16_t *input, tran_low_t *output, int stride);
|
||||
RTCD_EXTERN void (*vp9_fdct32x32)(const int16_t *input, tran_low_t *output, int stride);
|
||||
|
||||
void vp9_fdct32x32_1_c(const int16_t *input, tran_low_t *output, int stride);
|
||||
void vp9_fdct32x32_1_sse2(const int16_t *input, tran_low_t *output, int stride);
|
||||
@@ -255,7 +260,8 @@ void vp9_fdct32x32_1_sse2(const int16_t *input, tran_low_t *output, int stride);
|
||||
|
||||
void vp9_fdct32x32_rd_c(const int16_t *input, tran_low_t *output, int stride);
|
||||
void vp9_fdct32x32_rd_sse2(const int16_t *input, tran_low_t *output, int stride);
|
||||
#define vp9_fdct32x32_rd vp9_fdct32x32_rd_sse2
|
||||
void vp9_fdct32x32_rd_avx2(const int16_t *input, tran_low_t *output, int stride);
|
||||
RTCD_EXTERN void (*vp9_fdct32x32_rd)(const int16_t *input, tran_low_t *output, int stride);
|
||||
|
||||
void vp9_fdct4x4_c(const int16_t *input, tran_low_t *output, int stride);
|
||||
void vp9_fdct4x4_sse2(const int16_t *input, tran_low_t *output, int stride);
|
||||
@@ -402,7 +408,8 @@ void vp9_iwht4x4_1_add_c(const tran_low_t *input, uint8_t *dest, int dest_stride
|
||||
|
||||
void vp9_lpf_horizontal_16_c(uint8_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh, int count);
|
||||
void vp9_lpf_horizontal_16_sse2(uint8_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh, int count);
|
||||
#define vp9_lpf_horizontal_16 vp9_lpf_horizontal_16_sse2
|
||||
void vp9_lpf_horizontal_16_avx2(uint8_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh, int count);
|
||||
RTCD_EXTERN void (*vp9_lpf_horizontal_16)(uint8_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh, int count);
|
||||
|
||||
void vp9_lpf_horizontal_4_c(uint8_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh, int count);
|
||||
void vp9_lpf_horizontal_4_mmx(uint8_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh, int count);
|
||||
@@ -493,6 +500,7 @@ RTCD_EXTERN unsigned int (*vp9_sub_pixel_avg_variance32x16)(const uint8_t *src_p
|
||||
unsigned int vp9_sub_pixel_avg_variance32x32_c(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);
|
||||
unsigned int vp9_sub_pixel_avg_variance32x32_sse2(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);
|
||||
unsigned int vp9_sub_pixel_avg_variance32x32_ssse3(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);
|
||||
unsigned int vp9_sub_pixel_avg_variance32x32_avx2(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);
|
||||
RTCD_EXTERN unsigned int (*vp9_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);
|
||||
|
||||
unsigned int vp9_sub_pixel_avg_variance32x64_c(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);
|
||||
@@ -518,6 +526,7 @@ RTCD_EXTERN unsigned int (*vp9_sub_pixel_avg_variance64x32)(const uint8_t *src_p
|
||||
unsigned int vp9_sub_pixel_avg_variance64x64_c(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);
|
||||
unsigned int vp9_sub_pixel_avg_variance64x64_sse2(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);
|
||||
unsigned int vp9_sub_pixel_avg_variance64x64_ssse3(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);
|
||||
unsigned int vp9_sub_pixel_avg_variance64x64_avx2(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);
|
||||
RTCD_EXTERN unsigned int (*vp9_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);
|
||||
|
||||
unsigned int vp9_sub_pixel_avg_variance8x16_c(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);
|
||||
@@ -558,6 +567,7 @@ RTCD_EXTERN unsigned int (*vp9_sub_pixel_variance32x16)(const uint8_t *src_ptr,
|
||||
unsigned int vp9_sub_pixel_variance32x32_c(const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
unsigned int vp9_sub_pixel_variance32x32_sse2(const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
unsigned int vp9_sub_pixel_variance32x32_ssse3(const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
unsigned int vp9_sub_pixel_variance32x32_avx2(const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
RTCD_EXTERN unsigned int (*vp9_sub_pixel_variance32x32)(const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
|
||||
unsigned int vp9_sub_pixel_variance32x64_c(const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
@@ -583,6 +593,7 @@ RTCD_EXTERN unsigned int (*vp9_sub_pixel_variance64x32)(const uint8_t *src_ptr,
|
||||
unsigned int vp9_sub_pixel_variance64x64_c(const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
unsigned int vp9_sub_pixel_variance64x64_sse2(const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
unsigned int vp9_sub_pixel_variance64x64_ssse3(const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
unsigned int vp9_sub_pixel_variance64x64_avx2(const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
RTCD_EXTERN unsigned int (*vp9_sub_pixel_variance64x64)(const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
|
||||
unsigned int vp9_sub_pixel_variance8x16_c(const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
@@ -654,8 +665,11 @@ static void setup_rtcd_internal(void)
|
||||
|
||||
(void)flags;
|
||||
|
||||
vp9_block_error = vp9_block_error_sse2;
|
||||
if (flags & HAS_AVX2) vp9_block_error = vp9_block_error_avx2;
|
||||
vp9_convolve8 = vp9_convolve8_sse2;
|
||||
if (flags & HAS_SSSE3) vp9_convolve8 = vp9_convolve8_ssse3;
|
||||
if (flags & HAS_AVX2) vp9_convolve8 = vp9_convolve8_avx2;
|
||||
vp9_convolve8_avg = vp9_convolve8_avg_sse2;
|
||||
if (flags & HAS_SSSE3) vp9_convolve8_avg = vp9_convolve8_avg_ssse3;
|
||||
vp9_convolve8_avg_horiz = vp9_convolve8_avg_horiz_sse2;
|
||||
@@ -664,8 +678,10 @@ static void setup_rtcd_internal(void)
|
||||
if (flags & HAS_SSSE3) vp9_convolve8_avg_vert = vp9_convolve8_avg_vert_ssse3;
|
||||
vp9_convolve8_horiz = vp9_convolve8_horiz_sse2;
|
||||
if (flags & HAS_SSSE3) vp9_convolve8_horiz = vp9_convolve8_horiz_ssse3;
|
||||
if (flags & HAS_AVX2) vp9_convolve8_horiz = vp9_convolve8_horiz_avx2;
|
||||
vp9_convolve8_vert = vp9_convolve8_vert_sse2;
|
||||
if (flags & HAS_SSSE3) vp9_convolve8_vert = vp9_convolve8_vert_ssse3;
|
||||
if (flags & HAS_AVX2) vp9_convolve8_vert = vp9_convolve8_vert_avx2;
|
||||
vp9_d153_predictor_16x16 = vp9_d153_predictor_16x16_c;
|
||||
if (flags & HAS_SSSE3) vp9_d153_predictor_16x16 = vp9_d153_predictor_16x16_ssse3;
|
||||
vp9_d153_predictor_4x4 = vp9_d153_predictor_4x4_c;
|
||||
@@ -696,6 +712,10 @@ static void setup_rtcd_internal(void)
|
||||
if (flags & HAS_SSSE3) vp9_d63_predictor_4x4 = vp9_d63_predictor_4x4_ssse3;
|
||||
vp9_d63_predictor_8x8 = vp9_d63_predictor_8x8_c;
|
||||
if (flags & HAS_SSSE3) vp9_d63_predictor_8x8 = vp9_d63_predictor_8x8_ssse3;
|
||||
vp9_fdct32x32 = vp9_fdct32x32_sse2;
|
||||
if (flags & HAS_AVX2) vp9_fdct32x32 = vp9_fdct32x32_avx2;
|
||||
vp9_fdct32x32_rd = vp9_fdct32x32_rd_sse2;
|
||||
if (flags & HAS_AVX2) vp9_fdct32x32_rd = vp9_fdct32x32_rd_avx2;
|
||||
vp9_fdct8x8 = vp9_fdct8x8_sse2;
|
||||
if (flags & HAS_SSSE3) vp9_fdct8x8 = vp9_fdct8x8_ssse3;
|
||||
vp9_fdct8x8_quant = vp9_fdct8x8_quant_sse2;
|
||||
@@ -717,6 +737,8 @@ static void setup_rtcd_internal(void)
|
||||
if (flags & HAS_SSSE3) vp9_idct8x8_12_add = vp9_idct8x8_12_add_ssse3;
|
||||
vp9_idct8x8_64_add = vp9_idct8x8_64_add_sse2;
|
||||
if (flags & HAS_SSSE3) vp9_idct8x8_64_add = vp9_idct8x8_64_add_ssse3;
|
||||
vp9_lpf_horizontal_16 = vp9_lpf_horizontal_16_sse2;
|
||||
if (flags & HAS_AVX2) vp9_lpf_horizontal_16 = vp9_lpf_horizontal_16_avx2;
|
||||
vp9_quantize_b = vp9_quantize_b_sse2;
|
||||
if (flags & HAS_SSSE3) vp9_quantize_b = vp9_quantize_b_ssse3;
|
||||
vp9_quantize_b_32x32 = vp9_quantize_b_32x32_c;
|
||||
@@ -735,6 +757,7 @@ static void setup_rtcd_internal(void)
|
||||
if (flags & HAS_SSSE3) vp9_sub_pixel_avg_variance32x16 = vp9_sub_pixel_avg_variance32x16_ssse3;
|
||||
vp9_sub_pixel_avg_variance32x32 = vp9_sub_pixel_avg_variance32x32_sse2;
|
||||
if (flags & HAS_SSSE3) vp9_sub_pixel_avg_variance32x32 = vp9_sub_pixel_avg_variance32x32_ssse3;
|
||||
if (flags & HAS_AVX2) vp9_sub_pixel_avg_variance32x32 = vp9_sub_pixel_avg_variance32x32_avx2;
|
||||
vp9_sub_pixel_avg_variance32x64 = vp9_sub_pixel_avg_variance32x64_sse2;
|
||||
if (flags & HAS_SSSE3) vp9_sub_pixel_avg_variance32x64 = vp9_sub_pixel_avg_variance32x64_ssse3;
|
||||
vp9_sub_pixel_avg_variance4x4 = vp9_sub_pixel_avg_variance4x4_sse;
|
||||
@@ -745,6 +768,7 @@ static void setup_rtcd_internal(void)
|
||||
if (flags & HAS_SSSE3) vp9_sub_pixel_avg_variance64x32 = vp9_sub_pixel_avg_variance64x32_ssse3;
|
||||
vp9_sub_pixel_avg_variance64x64 = vp9_sub_pixel_avg_variance64x64_sse2;
|
||||
if (flags & HAS_SSSE3) vp9_sub_pixel_avg_variance64x64 = vp9_sub_pixel_avg_variance64x64_ssse3;
|
||||
if (flags & HAS_AVX2) vp9_sub_pixel_avg_variance64x64 = vp9_sub_pixel_avg_variance64x64_avx2;
|
||||
vp9_sub_pixel_avg_variance8x16 = vp9_sub_pixel_avg_variance8x16_sse2;
|
||||
if (flags & HAS_SSSE3) vp9_sub_pixel_avg_variance8x16 = vp9_sub_pixel_avg_variance8x16_ssse3;
|
||||
vp9_sub_pixel_avg_variance8x4 = vp9_sub_pixel_avg_variance8x4_sse2;
|
||||
@@ -761,6 +785,7 @@ static void setup_rtcd_internal(void)
|
||||
if (flags & HAS_SSSE3) vp9_sub_pixel_variance32x16 = vp9_sub_pixel_variance32x16_ssse3;
|
||||
vp9_sub_pixel_variance32x32 = vp9_sub_pixel_variance32x32_sse2;
|
||||
if (flags & HAS_SSSE3) vp9_sub_pixel_variance32x32 = vp9_sub_pixel_variance32x32_ssse3;
|
||||
if (flags & HAS_AVX2) vp9_sub_pixel_variance32x32 = vp9_sub_pixel_variance32x32_avx2;
|
||||
vp9_sub_pixel_variance32x64 = vp9_sub_pixel_variance32x64_sse2;
|
||||
if (flags & HAS_SSSE3) vp9_sub_pixel_variance32x64 = vp9_sub_pixel_variance32x64_ssse3;
|
||||
vp9_sub_pixel_variance4x4 = vp9_sub_pixel_variance4x4_sse;
|
||||
@@ -771,6 +796,7 @@ static void setup_rtcd_internal(void)
|
||||
if (flags & HAS_SSSE3) vp9_sub_pixel_variance64x32 = vp9_sub_pixel_variance64x32_ssse3;
|
||||
vp9_sub_pixel_variance64x64 = vp9_sub_pixel_variance64x64_sse2;
|
||||
if (flags & HAS_SSSE3) vp9_sub_pixel_variance64x64 = vp9_sub_pixel_variance64x64_ssse3;
|
||||
if (flags & HAS_AVX2) vp9_sub_pixel_variance64x64 = vp9_sub_pixel_variance64x64_avx2;
|
||||
vp9_sub_pixel_variance8x16 = vp9_sub_pixel_variance8x16_sse2;
|
||||
if (flags & HAS_SSSE3) vp9_sub_pixel_variance8x16 = vp9_sub_pixel_variance8x16_ssse3;
|
||||
vp9_sub_pixel_variance8x4 = vp9_sub_pixel_variance8x4_sse2;
|
||||
|
||||
@@ -39,6 +39,7 @@ unsigned int vp9_avg_8x8_sse2(const uint8_t *, int p);
|
||||
|
||||
int64_t vp9_block_error_c(const tran_low_t *coeff, const tran_low_t *dqcoeff, intptr_t block_size, int64_t *ssz);
|
||||
int64_t vp9_block_error_sse2(const tran_low_t *coeff, const tran_low_t *dqcoeff, intptr_t block_size, int64_t *ssz);
|
||||
int64_t vp9_block_error_avx2(const tran_low_t *coeff, const tran_low_t *dqcoeff, intptr_t block_size, int64_t *ssz);
|
||||
RTCD_EXTERN int64_t (*vp9_block_error)(const tran_low_t *coeff, const tran_low_t *dqcoeff, intptr_t block_size, int64_t *ssz);
|
||||
|
||||
int64_t vp9_block_error_fp_c(const int16_t *coeff, const int16_t *dqcoeff, int block_size);
|
||||
@@ -48,6 +49,7 @@ int64_t vp9_block_error_fp_sse2(const int16_t *coeff, const int16_t *dqcoeff, in
|
||||
void vp9_convolve8_c(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);
|
||||
void vp9_convolve8_sse2(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);
|
||||
void vp9_convolve8_ssse3(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);
|
||||
void vp9_convolve8_avx2(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);
|
||||
RTCD_EXTERN 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);
|
||||
|
||||
void vp9_convolve8_avg_c(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);
|
||||
@@ -68,11 +70,13 @@ RTCD_EXTERN void (*vp9_convolve8_avg_vert)(const uint8_t *src, ptrdiff_t src_str
|
||||
void vp9_convolve8_horiz_c(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);
|
||||
void vp9_convolve8_horiz_sse2(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);
|
||||
void vp9_convolve8_horiz_ssse3(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);
|
||||
void vp9_convolve8_horiz_avx2(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);
|
||||
RTCD_EXTERN 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);
|
||||
|
||||
void vp9_convolve8_vert_c(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);
|
||||
void vp9_convolve8_vert_sse2(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);
|
||||
void vp9_convolve8_vert_ssse3(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);
|
||||
void vp9_convolve8_vert_avx2(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);
|
||||
RTCD_EXTERN 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);
|
||||
|
||||
void vp9_convolve_avg_c(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);
|
||||
@@ -247,6 +251,7 @@ void vp9_fdct16x16_1_sse2(const int16_t *input, tran_low_t *output, int stride);
|
||||
|
||||
void vp9_fdct32x32_c(const int16_t *input, tran_low_t *output, int stride);
|
||||
void vp9_fdct32x32_sse2(const int16_t *input, tran_low_t *output, int stride);
|
||||
void vp9_fdct32x32_avx2(const int16_t *input, tran_low_t *output, int stride);
|
||||
RTCD_EXTERN void (*vp9_fdct32x32)(const int16_t *input, tran_low_t *output, int stride);
|
||||
|
||||
void vp9_fdct32x32_1_c(const int16_t *input, tran_low_t *output, int stride);
|
||||
@@ -255,6 +260,7 @@ void vp9_fdct32x32_1_sse2(const int16_t *input, tran_low_t *output, int stride);
|
||||
|
||||
void vp9_fdct32x32_rd_c(const int16_t *input, tran_low_t *output, int stride);
|
||||
void vp9_fdct32x32_rd_sse2(const int16_t *input, tran_low_t *output, int stride);
|
||||
void vp9_fdct32x32_rd_avx2(const int16_t *input, tran_low_t *output, int stride);
|
||||
RTCD_EXTERN void (*vp9_fdct32x32_rd)(const int16_t *input, tran_low_t *output, int stride);
|
||||
|
||||
void vp9_fdct4x4_c(const int16_t *input, tran_low_t *output, int stride);
|
||||
@@ -402,6 +408,7 @@ void vp9_iwht4x4_1_add_c(const tran_low_t *input, uint8_t *dest, int dest_stride
|
||||
|
||||
void vp9_lpf_horizontal_16_c(uint8_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh, int count);
|
||||
void vp9_lpf_horizontal_16_sse2(uint8_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh, int count);
|
||||
void vp9_lpf_horizontal_16_avx2(uint8_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh, int count);
|
||||
RTCD_EXTERN void (*vp9_lpf_horizontal_16)(uint8_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh, int count);
|
||||
|
||||
void vp9_lpf_horizontal_4_c(uint8_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh, int count);
|
||||
@@ -493,6 +500,7 @@ RTCD_EXTERN unsigned int (*vp9_sub_pixel_avg_variance32x16)(const uint8_t *src_p
|
||||
unsigned int vp9_sub_pixel_avg_variance32x32_c(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);
|
||||
unsigned int vp9_sub_pixel_avg_variance32x32_sse2(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);
|
||||
unsigned int vp9_sub_pixel_avg_variance32x32_ssse3(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);
|
||||
unsigned int vp9_sub_pixel_avg_variance32x32_avx2(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);
|
||||
RTCD_EXTERN unsigned int (*vp9_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);
|
||||
|
||||
unsigned int vp9_sub_pixel_avg_variance32x64_c(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);
|
||||
@@ -518,6 +526,7 @@ RTCD_EXTERN unsigned int (*vp9_sub_pixel_avg_variance64x32)(const uint8_t *src_p
|
||||
unsigned int vp9_sub_pixel_avg_variance64x64_c(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);
|
||||
unsigned int vp9_sub_pixel_avg_variance64x64_sse2(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);
|
||||
unsigned int vp9_sub_pixel_avg_variance64x64_ssse3(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);
|
||||
unsigned int vp9_sub_pixel_avg_variance64x64_avx2(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);
|
||||
RTCD_EXTERN unsigned int (*vp9_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);
|
||||
|
||||
unsigned int vp9_sub_pixel_avg_variance8x16_c(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);
|
||||
@@ -558,6 +567,7 @@ RTCD_EXTERN unsigned int (*vp9_sub_pixel_variance32x16)(const uint8_t *src_ptr,
|
||||
unsigned int vp9_sub_pixel_variance32x32_c(const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
unsigned int vp9_sub_pixel_variance32x32_sse2(const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
unsigned int vp9_sub_pixel_variance32x32_ssse3(const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
unsigned int vp9_sub_pixel_variance32x32_avx2(const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
RTCD_EXTERN unsigned int (*vp9_sub_pixel_variance32x32)(const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
|
||||
unsigned int vp9_sub_pixel_variance32x64_c(const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
@@ -583,6 +593,7 @@ RTCD_EXTERN unsigned int (*vp9_sub_pixel_variance64x32)(const uint8_t *src_ptr,
|
||||
unsigned int vp9_sub_pixel_variance64x64_c(const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
unsigned int vp9_sub_pixel_variance64x64_sse2(const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
unsigned int vp9_sub_pixel_variance64x64_ssse3(const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
unsigned int vp9_sub_pixel_variance64x64_avx2(const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
RTCD_EXTERN unsigned int (*vp9_sub_pixel_variance64x64)(const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
|
||||
unsigned int vp9_sub_pixel_variance8x16_c(const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
@@ -655,8 +666,10 @@ static void setup_rtcd_internal(void)
|
||||
(void)flags;
|
||||
|
||||
vp9_block_error = vp9_block_error_sse2;
|
||||
if (flags & HAS_AVX2) vp9_block_error = vp9_block_error_avx2;
|
||||
vp9_convolve8 = vp9_convolve8_sse2;
|
||||
if (flags & HAS_SSSE3) vp9_convolve8 = vp9_convolve8_ssse3;
|
||||
if (flags & HAS_AVX2) vp9_convolve8 = vp9_convolve8_avx2;
|
||||
vp9_convolve8_avg = vp9_convolve8_avg_sse2;
|
||||
if (flags & HAS_SSSE3) vp9_convolve8_avg = vp9_convolve8_avg_ssse3;
|
||||
vp9_convolve8_avg_horiz = vp9_convolve8_avg_horiz_sse2;
|
||||
@@ -665,8 +678,10 @@ static void setup_rtcd_internal(void)
|
||||
if (flags & HAS_SSSE3) vp9_convolve8_avg_vert = vp9_convolve8_avg_vert_ssse3;
|
||||
vp9_convolve8_horiz = vp9_convolve8_horiz_sse2;
|
||||
if (flags & HAS_SSSE3) vp9_convolve8_horiz = vp9_convolve8_horiz_ssse3;
|
||||
if (flags & HAS_AVX2) vp9_convolve8_horiz = vp9_convolve8_horiz_avx2;
|
||||
vp9_convolve8_vert = vp9_convolve8_vert_sse2;
|
||||
if (flags & HAS_SSSE3) vp9_convolve8_vert = vp9_convolve8_vert_ssse3;
|
||||
if (flags & HAS_AVX2) vp9_convolve8_vert = vp9_convolve8_vert_avx2;
|
||||
vp9_d153_predictor_16x16 = vp9_d153_predictor_16x16_c;
|
||||
if (flags & HAS_SSSE3) vp9_d153_predictor_16x16 = vp9_d153_predictor_16x16_ssse3;
|
||||
vp9_d153_predictor_4x4 = vp9_d153_predictor_4x4_c;
|
||||
@@ -698,7 +713,9 @@ static void setup_rtcd_internal(void)
|
||||
vp9_d63_predictor_8x8 = vp9_d63_predictor_8x8_c;
|
||||
if (flags & HAS_SSSE3) vp9_d63_predictor_8x8 = vp9_d63_predictor_8x8_ssse3;
|
||||
vp9_fdct32x32 = vp9_fdct32x32_sse2;
|
||||
if (flags & HAS_AVX2) vp9_fdct32x32 = vp9_fdct32x32_avx2;
|
||||
vp9_fdct32x32_rd = vp9_fdct32x32_rd_sse2;
|
||||
if (flags & HAS_AVX2) vp9_fdct32x32_rd = vp9_fdct32x32_rd_avx2;
|
||||
vp9_fdct8x8 = vp9_fdct8x8_sse2;
|
||||
if (flags & HAS_SSSE3) vp9_fdct8x8 = vp9_fdct8x8_ssse3;
|
||||
vp9_fdct8x8_quant = vp9_fdct8x8_quant_sse2;
|
||||
@@ -721,6 +738,7 @@ static void setup_rtcd_internal(void)
|
||||
vp9_idct8x8_64_add = vp9_idct8x8_64_add_sse2;
|
||||
if (flags & HAS_SSSE3) vp9_idct8x8_64_add = vp9_idct8x8_64_add_ssse3;
|
||||
vp9_lpf_horizontal_16 = vp9_lpf_horizontal_16_sse2;
|
||||
if (flags & HAS_AVX2) vp9_lpf_horizontal_16 = vp9_lpf_horizontal_16_avx2;
|
||||
vp9_quantize_b = vp9_quantize_b_sse2;
|
||||
if (flags & HAS_SSSE3) vp9_quantize_b = vp9_quantize_b_ssse3;
|
||||
vp9_quantize_b_32x32 = vp9_quantize_b_32x32_c;
|
||||
@@ -739,6 +757,7 @@ static void setup_rtcd_internal(void)
|
||||
if (flags & HAS_SSSE3) vp9_sub_pixel_avg_variance32x16 = vp9_sub_pixel_avg_variance32x16_ssse3;
|
||||
vp9_sub_pixel_avg_variance32x32 = vp9_sub_pixel_avg_variance32x32_sse2;
|
||||
if (flags & HAS_SSSE3) vp9_sub_pixel_avg_variance32x32 = vp9_sub_pixel_avg_variance32x32_ssse3;
|
||||
if (flags & HAS_AVX2) vp9_sub_pixel_avg_variance32x32 = vp9_sub_pixel_avg_variance32x32_avx2;
|
||||
vp9_sub_pixel_avg_variance32x64 = vp9_sub_pixel_avg_variance32x64_sse2;
|
||||
if (flags & HAS_SSSE3) vp9_sub_pixel_avg_variance32x64 = vp9_sub_pixel_avg_variance32x64_ssse3;
|
||||
vp9_sub_pixel_avg_variance4x4 = vp9_sub_pixel_avg_variance4x4_sse;
|
||||
@@ -749,6 +768,7 @@ static void setup_rtcd_internal(void)
|
||||
if (flags & HAS_SSSE3) vp9_sub_pixel_avg_variance64x32 = vp9_sub_pixel_avg_variance64x32_ssse3;
|
||||
vp9_sub_pixel_avg_variance64x64 = vp9_sub_pixel_avg_variance64x64_sse2;
|
||||
if (flags & HAS_SSSE3) vp9_sub_pixel_avg_variance64x64 = vp9_sub_pixel_avg_variance64x64_ssse3;
|
||||
if (flags & HAS_AVX2) vp9_sub_pixel_avg_variance64x64 = vp9_sub_pixel_avg_variance64x64_avx2;
|
||||
vp9_sub_pixel_avg_variance8x16 = vp9_sub_pixel_avg_variance8x16_sse2;
|
||||
if (flags & HAS_SSSE3) vp9_sub_pixel_avg_variance8x16 = vp9_sub_pixel_avg_variance8x16_ssse3;
|
||||
vp9_sub_pixel_avg_variance8x4 = vp9_sub_pixel_avg_variance8x4_sse2;
|
||||
@@ -765,6 +785,7 @@ static void setup_rtcd_internal(void)
|
||||
if (flags & HAS_SSSE3) vp9_sub_pixel_variance32x16 = vp9_sub_pixel_variance32x16_ssse3;
|
||||
vp9_sub_pixel_variance32x32 = vp9_sub_pixel_variance32x32_sse2;
|
||||
if (flags & HAS_SSSE3) vp9_sub_pixel_variance32x32 = vp9_sub_pixel_variance32x32_ssse3;
|
||||
if (flags & HAS_AVX2) vp9_sub_pixel_variance32x32 = vp9_sub_pixel_variance32x32_avx2;
|
||||
vp9_sub_pixel_variance32x64 = vp9_sub_pixel_variance32x64_sse2;
|
||||
if (flags & HAS_SSSE3) vp9_sub_pixel_variance32x64 = vp9_sub_pixel_variance32x64_ssse3;
|
||||
vp9_sub_pixel_variance4x4 = vp9_sub_pixel_variance4x4_sse;
|
||||
@@ -775,6 +796,7 @@ static void setup_rtcd_internal(void)
|
||||
if (flags & HAS_SSSE3) vp9_sub_pixel_variance64x32 = vp9_sub_pixel_variance64x32_ssse3;
|
||||
vp9_sub_pixel_variance64x64 = vp9_sub_pixel_variance64x64_sse2;
|
||||
if (flags & HAS_SSSE3) vp9_sub_pixel_variance64x64 = vp9_sub_pixel_variance64x64_ssse3;
|
||||
if (flags & HAS_AVX2) vp9_sub_pixel_variance64x64 = vp9_sub_pixel_variance64x64_avx2;
|
||||
vp9_sub_pixel_variance8x16 = vp9_sub_pixel_variance8x16_sse2;
|
||||
if (flags & HAS_SSSE3) vp9_sub_pixel_variance8x16 = vp9_sub_pixel_variance8x16_ssse3;
|
||||
vp9_sub_pixel_variance8x4 = vp9_sub_pixel_variance8x4_sse2;
|
||||
|
||||
@@ -92,6 +92,6 @@
|
||||
#define CONFIG_SPATIAL_SVC 0
|
||||
#define CONFIG_FP_MB_STATS 0
|
||||
#define CONFIG_EMULATE_HARDWARE 0
|
||||
#define DECODE_WIDTH_LIMIT 4000
|
||||
#define DECODE_HEIGHT_LIMIT 3000
|
||||
#define DECODE_WIDTH_LIMIT 8192
|
||||
#define DECODE_HEIGHT_LIMIT 4608
|
||||
#endif /* VPX_CONFIG_H */
|
||||
|
||||
@@ -92,6 +92,6 @@
|
||||
#define CONFIG_SPATIAL_SVC 0
|
||||
#define CONFIG_FP_MB_STATS 0
|
||||
#define CONFIG_EMULATE_HARDWARE 0
|
||||
#define DECODE_WIDTH_LIMIT 4000
|
||||
#define DECODE_HEIGHT_LIMIT 3000
|
||||
#define DECODE_WIDTH_LIMIT 8192
|
||||
#define DECODE_HEIGHT_LIMIT 4608
|
||||
#endif /* VPX_CONFIG_H */
|
||||
|
||||
@@ -92,6 +92,6 @@
|
||||
#define CONFIG_SPATIAL_SVC 0
|
||||
#define CONFIG_FP_MB_STATS 0
|
||||
#define CONFIG_EMULATE_HARDWARE 0
|
||||
#define DECODE_WIDTH_LIMIT 4000
|
||||
#define DECODE_HEIGHT_LIMIT 3000
|
||||
#define DECODE_WIDTH_LIMIT 8192
|
||||
#define DECODE_HEIGHT_LIMIT 4608
|
||||
#endif /* VPX_CONFIG_H */
|
||||
|
||||
@@ -17,7 +17,7 @@ HAVE_SSE3 equ 1
|
||||
HAVE_SSSE3 equ 1
|
||||
HAVE_SSE4_1 equ 1
|
||||
HAVE_AVX equ 1
|
||||
HAVE_AVX2 equ 0
|
||||
HAVE_AVX2 equ 1
|
||||
HAVE_VPX_PORTS equ 1
|
||||
HAVE_STDINT_H equ 1
|
||||
HAVE_PTHREAD_H equ 1
|
||||
|
||||
@@ -29,7 +29,7 @@
|
||||
#define HAVE_SSSE3 1
|
||||
#define HAVE_SSE4_1 1
|
||||
#define HAVE_AVX 1
|
||||
#define HAVE_AVX2 0
|
||||
#define HAVE_AVX2 1
|
||||
#define HAVE_VPX_PORTS 1
|
||||
#define HAVE_STDINT_H 1
|
||||
#define HAVE_PTHREAD_H 1
|
||||
@@ -92,6 +92,6 @@
|
||||
#define CONFIG_SPATIAL_SVC 0
|
||||
#define CONFIG_FP_MB_STATS 0
|
||||
#define CONFIG_EMULATE_HARDWARE 0
|
||||
#define DECODE_WIDTH_LIMIT 4000
|
||||
#define DECODE_HEIGHT_LIMIT 3000
|
||||
#define DECODE_WIDTH_LIMIT 8192
|
||||
#define DECODE_HEIGHT_LIMIT 4608
|
||||
#endif /* VPX_CONFIG_H */
|
||||
|
||||
@@ -93,6 +93,6 @@
|
||||
#define CONFIG_SPATIAL_SVC 0
|
||||
#define CONFIG_FP_MB_STATS 0
|
||||
#define CONFIG_EMULATE_HARDWARE 0
|
||||
#define DECODE_WIDTH_LIMIT 4000
|
||||
#define DECODE_HEIGHT_LIMIT 3000
|
||||
#define DECODE_WIDTH_LIMIT 8192
|
||||
#define DECODE_HEIGHT_LIMIT 4608
|
||||
#endif /* VPX_CONFIG_H */
|
||||
|
||||
@@ -28,8 +28,8 @@
|
||||
#define HAVE_SSE3 1
|
||||
#define HAVE_SSSE3 1
|
||||
#define HAVE_SSE4_1 1
|
||||
#define HAVE_AVX 0
|
||||
#define HAVE_AVX2 0
|
||||
#define HAVE_AVX 1
|
||||
#define HAVE_AVX2 1
|
||||
#define HAVE_VPX_PORTS 1
|
||||
#define HAVE_STDINT_H 0
|
||||
#define HAVE_PTHREAD_H 0
|
||||
@@ -92,6 +92,6 @@
|
||||
#define CONFIG_SPATIAL_SVC 0
|
||||
#define CONFIG_FP_MB_STATS 0
|
||||
#define CONFIG_EMULATE_HARDWARE 0
|
||||
#define DECODE_WIDTH_LIMIT 4000
|
||||
#define DECODE_HEIGHT_LIMIT 3000
|
||||
#define DECODE_WIDTH_LIMIT 8192
|
||||
#define DECODE_HEIGHT_LIMIT 4608
|
||||
#endif /* VPX_CONFIG_H */
|
||||
|
||||
@@ -92,6 +92,6 @@
|
||||
#define CONFIG_SPATIAL_SVC 0
|
||||
#define CONFIG_FP_MB_STATS 0
|
||||
#define CONFIG_EMULATE_HARDWARE 0
|
||||
#define DECODE_WIDTH_LIMIT 4000
|
||||
#define DECODE_HEIGHT_LIMIT 3000
|
||||
#define DECODE_WIDTH_LIMIT 8192
|
||||
#define DECODE_HEIGHT_LIMIT 4608
|
||||
#endif /* VPX_CONFIG_H */
|
||||
|
||||
@@ -17,7 +17,7 @@ HAVE_SSE3 equ 1
|
||||
HAVE_SSSE3 equ 1
|
||||
HAVE_SSE4_1 equ 1
|
||||
HAVE_AVX equ 1
|
||||
HAVE_AVX2 equ 0
|
||||
HAVE_AVX2 equ 1
|
||||
HAVE_VPX_PORTS equ 1
|
||||
HAVE_STDINT_H equ 1
|
||||
HAVE_PTHREAD_H equ 1
|
||||
|
||||
@@ -29,7 +29,7 @@
|
||||
#define HAVE_SSSE3 1
|
||||
#define HAVE_SSE4_1 1
|
||||
#define HAVE_AVX 1
|
||||
#define HAVE_AVX2 0
|
||||
#define HAVE_AVX2 1
|
||||
#define HAVE_VPX_PORTS 1
|
||||
#define HAVE_STDINT_H 1
|
||||
#define HAVE_PTHREAD_H 1
|
||||
@@ -92,6 +92,6 @@
|
||||
#define CONFIG_SPATIAL_SVC 0
|
||||
#define CONFIG_FP_MB_STATS 0
|
||||
#define CONFIG_EMULATE_HARDWARE 0
|
||||
#define DECODE_WIDTH_LIMIT 4000
|
||||
#define DECODE_HEIGHT_LIMIT 3000
|
||||
#define DECODE_WIDTH_LIMIT 8192
|
||||
#define DECODE_HEIGHT_LIMIT 4608
|
||||
#endif /* VPX_CONFIG_H */
|
||||
|
||||
@@ -93,6 +93,6 @@
|
||||
#define CONFIG_SPATIAL_SVC 0
|
||||
#define CONFIG_FP_MB_STATS 0
|
||||
#define CONFIG_EMULATE_HARDWARE 0
|
||||
#define DECODE_WIDTH_LIMIT 4000
|
||||
#define DECODE_HEIGHT_LIMIT 3000
|
||||
#define DECODE_WIDTH_LIMIT 8192
|
||||
#define DECODE_HEIGHT_LIMIT 4608
|
||||
#endif /* VPX_CONFIG_H */
|
||||
|
||||
@@ -28,8 +28,8 @@
|
||||
#define HAVE_SSE3 1
|
||||
#define HAVE_SSSE3 1
|
||||
#define HAVE_SSE4_1 1
|
||||
#define HAVE_AVX 0
|
||||
#define HAVE_AVX2 0
|
||||
#define HAVE_AVX 1
|
||||
#define HAVE_AVX2 1
|
||||
#define HAVE_VPX_PORTS 1
|
||||
#define HAVE_STDINT_H 0
|
||||
#define HAVE_PTHREAD_H 0
|
||||
@@ -92,6 +92,6 @@
|
||||
#define CONFIG_SPATIAL_SVC 0
|
||||
#define CONFIG_FP_MB_STATS 0
|
||||
#define CONFIG_EMULATE_HARDWARE 0
|
||||
#define DECODE_WIDTH_LIMIT 4000
|
||||
#define DECODE_HEIGHT_LIMIT 3000
|
||||
#define DECODE_WIDTH_LIMIT 8192
|
||||
#define DECODE_HEIGHT_LIMIT 4608
|
||||
#endif /* VPX_CONFIG_H */
|
||||
|
||||
@@ -11,6 +11,9 @@
|
||||
#define RTCD_C
|
||||
#include "./vpx_dsp_rtcd.h"
|
||||
#include "vpx_ports/vpx_once.h"
|
||||
#ifdef _MSC_VER
|
||||
#include <intrin.h>
|
||||
#endif
|
||||
|
||||
void vpx_dsp_rtcd() {
|
||||
once(setup_rtcd_internal);
|
||||
|
||||
@@ -23,6 +23,7 @@ void vpx_comp_avg_pred_c(uint8_t *comp_pred, const uint8_t *pred, int width, int
|
||||
|
||||
void vpx_get16x16var_c(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, int *sum);
|
||||
void vpx_get16x16var_sse2(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, int *sum);
|
||||
void vpx_get16x16var_avx2(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, int *sum);
|
||||
RTCD_EXTERN void (*vpx_get16x16var)(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, int *sum);
|
||||
|
||||
unsigned int vpx_get4x4sse_cs_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride);
|
||||
@@ -41,6 +42,7 @@ RTCD_EXTERN unsigned int (*vpx_get_mb_ss)(const int16_t *);
|
||||
unsigned int vpx_mse16x16_c(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int recon_stride, unsigned int *sse);
|
||||
unsigned int vpx_mse16x16_mmx(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int recon_stride, unsigned int *sse);
|
||||
unsigned int vpx_mse16x16_sse2(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int recon_stride, unsigned int *sse);
|
||||
unsigned int vpx_mse16x16_avx2(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int recon_stride, unsigned int *sse);
|
||||
RTCD_EXTERN unsigned int (*vpx_mse16x16)(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int recon_stride, unsigned int *sse);
|
||||
|
||||
unsigned int vpx_mse16x8_c(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int recon_stride, unsigned int *sse);
|
||||
@@ -103,34 +105,41 @@ void vpx_sad16x8x8_sse4_1(const uint8_t *src_ptr, int src_stride, const uint8_t
|
||||
RTCD_EXTERN void (*vpx_sad16x8x8)(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, uint32_t *sad_array);
|
||||
|
||||
unsigned int vpx_sad32x16_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
#define vpx_sad32x16 vpx_sad32x16_c
|
||||
unsigned int vpx_sad32x16_avx2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
RTCD_EXTERN unsigned int (*vpx_sad32x16)(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
|
||||
unsigned int vpx_sad32x16_avg_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
#define vpx_sad32x16_avg vpx_sad32x16_avg_c
|
||||
unsigned int vpx_sad32x16_avg_avx2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
RTCD_EXTERN unsigned int (*vpx_sad32x16_avg)(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
|
||||
void vpx_sad32x16x4d_c(const uint8_t *src_ptr, int src_stride, const uint8_t * const ref_ptr[], int ref_stride, uint32_t *sad_array);
|
||||
#define vpx_sad32x16x4d vpx_sad32x16x4d_c
|
||||
|
||||
unsigned int vpx_sad32x32_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
#define vpx_sad32x32 vpx_sad32x32_c
|
||||
unsigned int vpx_sad32x32_avx2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
RTCD_EXTERN unsigned int (*vpx_sad32x32)(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
|
||||
unsigned int vpx_sad32x32_avg_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
#define vpx_sad32x32_avg vpx_sad32x32_avg_c
|
||||
unsigned int vpx_sad32x32_avg_avx2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
RTCD_EXTERN unsigned int (*vpx_sad32x32_avg)(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
|
||||
void vpx_sad32x32x3_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, uint32_t *sad_array);
|
||||
#define vpx_sad32x32x3 vpx_sad32x32x3_c
|
||||
|
||||
void vpx_sad32x32x4d_c(const uint8_t *src_ptr, int src_stride, const uint8_t * const ref_ptr[], int ref_stride, uint32_t *sad_array);
|
||||
#define vpx_sad32x32x4d vpx_sad32x32x4d_c
|
||||
void vpx_sad32x32x4d_avx2(const uint8_t *src_ptr, int src_stride, const uint8_t * const ref_ptr[], int ref_stride, uint32_t *sad_array);
|
||||
RTCD_EXTERN void (*vpx_sad32x32x4d)(const uint8_t *src_ptr, int src_stride, const uint8_t * const ref_ptr[], int ref_stride, uint32_t *sad_array);
|
||||
|
||||
void vpx_sad32x32x8_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, uint32_t *sad_array);
|
||||
#define vpx_sad32x32x8 vpx_sad32x32x8_c
|
||||
|
||||
unsigned int vpx_sad32x64_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
#define vpx_sad32x64 vpx_sad32x64_c
|
||||
unsigned int vpx_sad32x64_avx2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
RTCD_EXTERN unsigned int (*vpx_sad32x64)(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
|
||||
unsigned int vpx_sad32x64_avg_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
#define vpx_sad32x64_avg vpx_sad32x64_avg_c
|
||||
unsigned int vpx_sad32x64_avg_avx2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
RTCD_EXTERN unsigned int (*vpx_sad32x64_avg)(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
|
||||
void vpx_sad32x64x4d_c(const uint8_t *src_ptr, int src_stride, const uint8_t * const ref_ptr[], int ref_stride, uint32_t *sad_array);
|
||||
#define vpx_sad32x64x4d vpx_sad32x64x4d_c
|
||||
@@ -166,25 +175,30 @@ void vpx_sad4x8x8_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_p
|
||||
#define vpx_sad4x8x8 vpx_sad4x8x8_c
|
||||
|
||||
unsigned int vpx_sad64x32_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
#define vpx_sad64x32 vpx_sad64x32_c
|
||||
unsigned int vpx_sad64x32_avx2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
RTCD_EXTERN unsigned int (*vpx_sad64x32)(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
|
||||
unsigned int vpx_sad64x32_avg_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
#define vpx_sad64x32_avg vpx_sad64x32_avg_c
|
||||
unsigned int vpx_sad64x32_avg_avx2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
RTCD_EXTERN unsigned int (*vpx_sad64x32_avg)(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
|
||||
void vpx_sad64x32x4d_c(const uint8_t *src_ptr, int src_stride, const uint8_t * const ref_ptr[], int ref_stride, uint32_t *sad_array);
|
||||
#define vpx_sad64x32x4d vpx_sad64x32x4d_c
|
||||
|
||||
unsigned int vpx_sad64x64_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
#define vpx_sad64x64 vpx_sad64x64_c
|
||||
unsigned int vpx_sad64x64_avx2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
RTCD_EXTERN unsigned int (*vpx_sad64x64)(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
|
||||
unsigned int vpx_sad64x64_avg_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
#define vpx_sad64x64_avg vpx_sad64x64_avg_c
|
||||
unsigned int vpx_sad64x64_avg_avx2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
RTCD_EXTERN unsigned int (*vpx_sad64x64_avg)(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
|
||||
void vpx_sad64x64x3_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, uint32_t *sad_array);
|
||||
#define vpx_sad64x64x3 vpx_sad64x64x3_c
|
||||
|
||||
void vpx_sad64x64x4d_c(const uint8_t *src_ptr, int src_stride, const uint8_t * const ref_ptr[], int ref_stride, uint32_t *sad_array);
|
||||
#define vpx_sad64x64x4d vpx_sad64x64x4d_c
|
||||
void vpx_sad64x64x4d_avx2(const uint8_t *src_ptr, int src_stride, const uint8_t * const ref_ptr[], int ref_stride, uint32_t *sad_array);
|
||||
RTCD_EXTERN void (*vpx_sad64x64x4d)(const uint8_t *src_ptr, int src_stride, const uint8_t * const ref_ptr[], int ref_stride, uint32_t *sad_array);
|
||||
|
||||
void vpx_sad64x64x8_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, uint32_t *sad_array);
|
||||
#define vpx_sad64x64x8 vpx_sad64x64x8_c
|
||||
@@ -240,6 +254,7 @@ RTCD_EXTERN void (*vpx_sad8x8x8)(const uint8_t *src_ptr, int src_stride, const u
|
||||
unsigned int vpx_variance16x16_c(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
unsigned int vpx_variance16x16_mmx(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
unsigned int vpx_variance16x16_sse2(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
unsigned int vpx_variance16x16_avx2(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
RTCD_EXTERN unsigned int (*vpx_variance16x16)(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
|
||||
unsigned int vpx_variance16x32_c(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
@@ -253,10 +268,12 @@ RTCD_EXTERN unsigned int (*vpx_variance16x8)(const uint8_t *src_ptr, int source_
|
||||
|
||||
unsigned int vpx_variance32x16_c(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
unsigned int vpx_variance32x16_sse2(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
unsigned int vpx_variance32x16_avx2(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
RTCD_EXTERN unsigned int (*vpx_variance32x16)(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
|
||||
unsigned int vpx_variance32x32_c(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
unsigned int vpx_variance32x32_sse2(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
unsigned int vpx_variance32x32_avx2(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
RTCD_EXTERN unsigned int (*vpx_variance32x32)(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
|
||||
unsigned int vpx_variance32x64_c(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
@@ -274,10 +291,12 @@ RTCD_EXTERN unsigned int (*vpx_variance4x8)(const uint8_t *src_ptr, int source_s
|
||||
|
||||
unsigned int vpx_variance64x32_c(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
unsigned int vpx_variance64x32_sse2(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
unsigned int vpx_variance64x32_avx2(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
RTCD_EXTERN unsigned int (*vpx_variance64x32)(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
|
||||
unsigned int vpx_variance64x64_c(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
unsigned int vpx_variance64x64_sse2(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
unsigned int vpx_variance64x64_avx2(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
RTCD_EXTERN unsigned int (*vpx_variance64x64)(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
|
||||
unsigned int vpx_variance8x16_c(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
@@ -306,6 +325,7 @@ static void setup_rtcd_internal(void)
|
||||
|
||||
vpx_get16x16var = vpx_get16x16var_c;
|
||||
if (flags & HAS_SSE2) vpx_get16x16var = vpx_get16x16var_sse2;
|
||||
if (flags & HAS_AVX2) vpx_get16x16var = vpx_get16x16var_avx2;
|
||||
vpx_get8x8var = vpx_get8x8var_c;
|
||||
if (flags & HAS_MMX) vpx_get8x8var = vpx_get8x8var_mmx;
|
||||
if (flags & HAS_SSE2) vpx_get8x8var = vpx_get8x8var_sse2;
|
||||
@@ -315,6 +335,7 @@ static void setup_rtcd_internal(void)
|
||||
vpx_mse16x16 = vpx_mse16x16_c;
|
||||
if (flags & HAS_MMX) vpx_mse16x16 = vpx_mse16x16_mmx;
|
||||
if (flags & HAS_SSE2) vpx_mse16x16 = vpx_mse16x16_sse2;
|
||||
if (flags & HAS_AVX2) vpx_mse16x16 = vpx_mse16x16_avx2;
|
||||
vpx_mse16x8 = vpx_mse16x8_c;
|
||||
if (flags & HAS_SSE2) vpx_mse16x8 = vpx_mse16x8_sse2;
|
||||
vpx_mse8x16 = vpx_mse8x16_c;
|
||||
@@ -335,12 +356,36 @@ static void setup_rtcd_internal(void)
|
||||
if (flags & HAS_SSSE3) vpx_sad16x8x3 = vpx_sad16x8x3_ssse3;
|
||||
vpx_sad16x8x8 = vpx_sad16x8x8_c;
|
||||
if (flags & HAS_SSE4_1) vpx_sad16x8x8 = vpx_sad16x8x8_sse4_1;
|
||||
vpx_sad32x16 = vpx_sad32x16_c;
|
||||
if (flags & HAS_AVX2) vpx_sad32x16 = vpx_sad32x16_avx2;
|
||||
vpx_sad32x16_avg = vpx_sad32x16_avg_c;
|
||||
if (flags & HAS_AVX2) vpx_sad32x16_avg = vpx_sad32x16_avg_avx2;
|
||||
vpx_sad32x32 = vpx_sad32x32_c;
|
||||
if (flags & HAS_AVX2) vpx_sad32x32 = vpx_sad32x32_avx2;
|
||||
vpx_sad32x32_avg = vpx_sad32x32_avg_c;
|
||||
if (flags & HAS_AVX2) vpx_sad32x32_avg = vpx_sad32x32_avg_avx2;
|
||||
vpx_sad32x32x4d = vpx_sad32x32x4d_c;
|
||||
if (flags & HAS_AVX2) vpx_sad32x32x4d = vpx_sad32x32x4d_avx2;
|
||||
vpx_sad32x64 = vpx_sad32x64_c;
|
||||
if (flags & HAS_AVX2) vpx_sad32x64 = vpx_sad32x64_avx2;
|
||||
vpx_sad32x64_avg = vpx_sad32x64_avg_c;
|
||||
if (flags & HAS_AVX2) vpx_sad32x64_avg = vpx_sad32x64_avg_avx2;
|
||||
vpx_sad4x4 = vpx_sad4x4_c;
|
||||
if (flags & HAS_MMX) vpx_sad4x4 = vpx_sad4x4_mmx;
|
||||
vpx_sad4x4x3 = vpx_sad4x4x3_c;
|
||||
if (flags & HAS_SSE3) vpx_sad4x4x3 = vpx_sad4x4x3_sse3;
|
||||
vpx_sad4x4x8 = vpx_sad4x4x8_c;
|
||||
if (flags & HAS_SSE4_1) vpx_sad4x4x8 = vpx_sad4x4x8_sse4_1;
|
||||
vpx_sad64x32 = vpx_sad64x32_c;
|
||||
if (flags & HAS_AVX2) vpx_sad64x32 = vpx_sad64x32_avx2;
|
||||
vpx_sad64x32_avg = vpx_sad64x32_avg_c;
|
||||
if (flags & HAS_AVX2) vpx_sad64x32_avg = vpx_sad64x32_avg_avx2;
|
||||
vpx_sad64x64 = vpx_sad64x64_c;
|
||||
if (flags & HAS_AVX2) vpx_sad64x64 = vpx_sad64x64_avx2;
|
||||
vpx_sad64x64_avg = vpx_sad64x64_avg_c;
|
||||
if (flags & HAS_AVX2) vpx_sad64x64_avg = vpx_sad64x64_avg_avx2;
|
||||
vpx_sad64x64x4d = vpx_sad64x64x4d_c;
|
||||
if (flags & HAS_AVX2) vpx_sad64x64x4d = vpx_sad64x64x4d_avx2;
|
||||
vpx_sad8x16 = vpx_sad8x16_c;
|
||||
if (flags & HAS_MMX) vpx_sad8x16 = vpx_sad8x16_mmx;
|
||||
vpx_sad8x16x3 = vpx_sad8x16x3_c;
|
||||
@@ -356,6 +401,7 @@ static void setup_rtcd_internal(void)
|
||||
vpx_variance16x16 = vpx_variance16x16_c;
|
||||
if (flags & HAS_MMX) vpx_variance16x16 = vpx_variance16x16_mmx;
|
||||
if (flags & HAS_SSE2) vpx_variance16x16 = vpx_variance16x16_sse2;
|
||||
if (flags & HAS_AVX2) vpx_variance16x16 = vpx_variance16x16_avx2;
|
||||
vpx_variance16x32 = vpx_variance16x32_c;
|
||||
if (flags & HAS_SSE2) vpx_variance16x32 = vpx_variance16x32_sse2;
|
||||
vpx_variance16x8 = vpx_variance16x8_c;
|
||||
@@ -363,8 +409,10 @@ static void setup_rtcd_internal(void)
|
||||
if (flags & HAS_SSE2) vpx_variance16x8 = vpx_variance16x8_sse2;
|
||||
vpx_variance32x16 = vpx_variance32x16_c;
|
||||
if (flags & HAS_SSE2) vpx_variance32x16 = vpx_variance32x16_sse2;
|
||||
if (flags & HAS_AVX2) vpx_variance32x16 = vpx_variance32x16_avx2;
|
||||
vpx_variance32x32 = vpx_variance32x32_c;
|
||||
if (flags & HAS_SSE2) vpx_variance32x32 = vpx_variance32x32_sse2;
|
||||
if (flags & HAS_AVX2) vpx_variance32x32 = vpx_variance32x32_avx2;
|
||||
vpx_variance32x64 = vpx_variance32x64_c;
|
||||
if (flags & HAS_SSE2) vpx_variance32x64 = vpx_variance32x64_sse2;
|
||||
vpx_variance4x4 = vpx_variance4x4_c;
|
||||
@@ -374,8 +422,10 @@ static void setup_rtcd_internal(void)
|
||||
if (flags & HAS_SSE2) vpx_variance4x8 = vpx_variance4x8_sse2;
|
||||
vpx_variance64x32 = vpx_variance64x32_c;
|
||||
if (flags & HAS_SSE2) vpx_variance64x32 = vpx_variance64x32_sse2;
|
||||
if (flags & HAS_AVX2) vpx_variance64x32 = vpx_variance64x32_avx2;
|
||||
vpx_variance64x64 = vpx_variance64x64_c;
|
||||
if (flags & HAS_SSE2) vpx_variance64x64 = vpx_variance64x64_sse2;
|
||||
if (flags & HAS_AVX2) vpx_variance64x64 = vpx_variance64x64_avx2;
|
||||
vpx_variance8x16 = vpx_variance8x16_c;
|
||||
if (flags & HAS_MMX) vpx_variance8x16 = vpx_variance8x16_mmx;
|
||||
if (flags & HAS_SSE2) vpx_variance8x16 = vpx_variance8x16_sse2;
|
||||
|
||||
@@ -23,6 +23,7 @@ void vpx_comp_avg_pred_c(uint8_t *comp_pred, const uint8_t *pred, int width, int
|
||||
|
||||
void vpx_get16x16var_c(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, int *sum);
|
||||
void vpx_get16x16var_sse2(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, int *sum);
|
||||
void vpx_get16x16var_avx2(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, int *sum);
|
||||
RTCD_EXTERN void (*vpx_get16x16var)(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, int *sum);
|
||||
|
||||
unsigned int vpx_get4x4sse_cs_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride);
|
||||
@@ -41,6 +42,7 @@ RTCD_EXTERN unsigned int (*vpx_get_mb_ss)(const int16_t *);
|
||||
unsigned int vpx_mse16x16_c(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int recon_stride, unsigned int *sse);
|
||||
unsigned int vpx_mse16x16_mmx(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int recon_stride, unsigned int *sse);
|
||||
unsigned int vpx_mse16x16_sse2(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int recon_stride, unsigned int *sse);
|
||||
unsigned int vpx_mse16x16_avx2(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int recon_stride, unsigned int *sse);
|
||||
RTCD_EXTERN unsigned int (*vpx_mse16x16)(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int recon_stride, unsigned int *sse);
|
||||
|
||||
unsigned int vpx_mse16x8_c(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int recon_stride, unsigned int *sse);
|
||||
@@ -113,10 +115,12 @@ RTCD_EXTERN void (*vpx_sad16x8x8)(const uint8_t *src_ptr, int src_stride, const
|
||||
|
||||
unsigned int vpx_sad32x16_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
unsigned int vpx_sad32x16_sse2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
unsigned int vpx_sad32x16_avx2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
RTCD_EXTERN unsigned int (*vpx_sad32x16)(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
|
||||
unsigned int vpx_sad32x16_avg_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
unsigned int vpx_sad32x16_avg_sse2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
unsigned int vpx_sad32x16_avg_avx2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
RTCD_EXTERN unsigned int (*vpx_sad32x16_avg)(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
|
||||
void vpx_sad32x16x4d_c(const uint8_t *src_ptr, int src_stride, const uint8_t * const ref_ptr[], int ref_stride, uint32_t *sad_array);
|
||||
@@ -125,10 +129,12 @@ RTCD_EXTERN void (*vpx_sad32x16x4d)(const uint8_t *src_ptr, int src_stride, cons
|
||||
|
||||
unsigned int vpx_sad32x32_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
unsigned int vpx_sad32x32_sse2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
unsigned int vpx_sad32x32_avx2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
RTCD_EXTERN unsigned int (*vpx_sad32x32)(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
|
||||
unsigned int vpx_sad32x32_avg_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
unsigned int vpx_sad32x32_avg_sse2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
unsigned int vpx_sad32x32_avg_avx2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
RTCD_EXTERN unsigned int (*vpx_sad32x32_avg)(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
|
||||
void vpx_sad32x32x3_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, uint32_t *sad_array);
|
||||
@@ -136,6 +142,7 @@ void vpx_sad32x32x3_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref
|
||||
|
||||
void vpx_sad32x32x4d_c(const uint8_t *src_ptr, int src_stride, const uint8_t * const ref_ptr[], int ref_stride, uint32_t *sad_array);
|
||||
void vpx_sad32x32x4d_sse2(const uint8_t *src_ptr, int src_stride, const uint8_t * const ref_ptr[], int ref_stride, uint32_t *sad_array);
|
||||
void vpx_sad32x32x4d_avx2(const uint8_t *src_ptr, int src_stride, const uint8_t * const ref_ptr[], int ref_stride, uint32_t *sad_array);
|
||||
RTCD_EXTERN void (*vpx_sad32x32x4d)(const uint8_t *src_ptr, int src_stride, const uint8_t * const ref_ptr[], int ref_stride, uint32_t *sad_array);
|
||||
|
||||
void vpx_sad32x32x8_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, uint32_t *sad_array);
|
||||
@@ -143,10 +150,12 @@ void vpx_sad32x32x8_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref
|
||||
|
||||
unsigned int vpx_sad32x64_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
unsigned int vpx_sad32x64_sse2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
unsigned int vpx_sad32x64_avx2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
RTCD_EXTERN unsigned int (*vpx_sad32x64)(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
|
||||
unsigned int vpx_sad32x64_avg_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
unsigned int vpx_sad32x64_avg_sse2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
unsigned int vpx_sad32x64_avg_avx2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
RTCD_EXTERN unsigned int (*vpx_sad32x64_avg)(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
|
||||
void vpx_sad32x64x4d_c(const uint8_t *src_ptr, int src_stride, const uint8_t * const ref_ptr[], int ref_stride, uint32_t *sad_array);
|
||||
@@ -191,10 +200,12 @@ void vpx_sad4x8x8_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_p
|
||||
|
||||
unsigned int vpx_sad64x32_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
unsigned int vpx_sad64x32_sse2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
unsigned int vpx_sad64x32_avx2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
RTCD_EXTERN unsigned int (*vpx_sad64x32)(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
|
||||
unsigned int vpx_sad64x32_avg_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
unsigned int vpx_sad64x32_avg_sse2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
unsigned int vpx_sad64x32_avg_avx2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
RTCD_EXTERN unsigned int (*vpx_sad64x32_avg)(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
|
||||
void vpx_sad64x32x4d_c(const uint8_t *src_ptr, int src_stride, const uint8_t * const ref_ptr[], int ref_stride, uint32_t *sad_array);
|
||||
@@ -203,10 +214,12 @@ RTCD_EXTERN void (*vpx_sad64x32x4d)(const uint8_t *src_ptr, int src_stride, cons
|
||||
|
||||
unsigned int vpx_sad64x64_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
unsigned int vpx_sad64x64_sse2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
unsigned int vpx_sad64x64_avx2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
RTCD_EXTERN unsigned int (*vpx_sad64x64)(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
|
||||
unsigned int vpx_sad64x64_avg_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
unsigned int vpx_sad64x64_avg_sse2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
unsigned int vpx_sad64x64_avg_avx2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
RTCD_EXTERN unsigned int (*vpx_sad64x64_avg)(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
|
||||
void vpx_sad64x64x3_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, uint32_t *sad_array);
|
||||
@@ -214,6 +227,7 @@ void vpx_sad64x64x3_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref
|
||||
|
||||
void vpx_sad64x64x4d_c(const uint8_t *src_ptr, int src_stride, const uint8_t * const ref_ptr[], int ref_stride, uint32_t *sad_array);
|
||||
void vpx_sad64x64x4d_sse2(const uint8_t *src_ptr, int src_stride, const uint8_t * const ref_ptr[], int ref_stride, uint32_t *sad_array);
|
||||
void vpx_sad64x64x4d_avx2(const uint8_t *src_ptr, int src_stride, const uint8_t * const ref_ptr[], int ref_stride, uint32_t *sad_array);
|
||||
RTCD_EXTERN void (*vpx_sad64x64x4d)(const uint8_t *src_ptr, int src_stride, const uint8_t * const ref_ptr[], int ref_stride, uint32_t *sad_array);
|
||||
|
||||
void vpx_sad64x64x8_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, uint32_t *sad_array);
|
||||
@@ -279,6 +293,7 @@ RTCD_EXTERN void (*vpx_sad8x8x8)(const uint8_t *src_ptr, int src_stride, const u
|
||||
unsigned int vpx_variance16x16_c(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
unsigned int vpx_variance16x16_mmx(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
unsigned int vpx_variance16x16_sse2(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
unsigned int vpx_variance16x16_avx2(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
RTCD_EXTERN unsigned int (*vpx_variance16x16)(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
|
||||
unsigned int vpx_variance16x32_c(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
@@ -292,10 +307,12 @@ RTCD_EXTERN unsigned int (*vpx_variance16x8)(const uint8_t *src_ptr, int source_
|
||||
|
||||
unsigned int vpx_variance32x16_c(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
unsigned int vpx_variance32x16_sse2(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
unsigned int vpx_variance32x16_avx2(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
RTCD_EXTERN unsigned int (*vpx_variance32x16)(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
|
||||
unsigned int vpx_variance32x32_c(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
unsigned int vpx_variance32x32_sse2(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
unsigned int vpx_variance32x32_avx2(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
RTCD_EXTERN unsigned int (*vpx_variance32x32)(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
|
||||
unsigned int vpx_variance32x64_c(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
@@ -313,10 +330,12 @@ RTCD_EXTERN unsigned int (*vpx_variance4x8)(const uint8_t *src_ptr, int source_s
|
||||
|
||||
unsigned int vpx_variance64x32_c(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
unsigned int vpx_variance64x32_sse2(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
unsigned int vpx_variance64x32_avx2(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
RTCD_EXTERN unsigned int (*vpx_variance64x32)(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
|
||||
unsigned int vpx_variance64x64_c(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
unsigned int vpx_variance64x64_sse2(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
unsigned int vpx_variance64x64_avx2(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
RTCD_EXTERN unsigned int (*vpx_variance64x64)(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
|
||||
unsigned int vpx_variance8x16_c(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
@@ -345,6 +364,7 @@ static void setup_rtcd_internal(void)
|
||||
|
||||
vpx_get16x16var = vpx_get16x16var_c;
|
||||
if (flags & HAS_SSE2) vpx_get16x16var = vpx_get16x16var_sse2;
|
||||
if (flags & HAS_AVX2) vpx_get16x16var = vpx_get16x16var_avx2;
|
||||
vpx_get8x8var = vpx_get8x8var_c;
|
||||
if (flags & HAS_MMX) vpx_get8x8var = vpx_get8x8var_mmx;
|
||||
if (flags & HAS_SSE2) vpx_get8x8var = vpx_get8x8var_sse2;
|
||||
@@ -354,6 +374,7 @@ static void setup_rtcd_internal(void)
|
||||
vpx_mse16x16 = vpx_mse16x16_c;
|
||||
if (flags & HAS_MMX) vpx_mse16x16 = vpx_mse16x16_mmx;
|
||||
if (flags & HAS_SSE2) vpx_mse16x16 = vpx_mse16x16_sse2;
|
||||
if (flags & HAS_AVX2) vpx_mse16x16 = vpx_mse16x16_avx2;
|
||||
vpx_mse16x8 = vpx_mse16x8_c;
|
||||
if (flags & HAS_SSE2) vpx_mse16x8 = vpx_mse16x8_sse2;
|
||||
vpx_mse8x16 = vpx_mse8x16_c;
|
||||
@@ -392,20 +413,27 @@ static void setup_rtcd_internal(void)
|
||||
if (flags & HAS_SSE4_1) vpx_sad16x8x8 = vpx_sad16x8x8_sse4_1;
|
||||
vpx_sad32x16 = vpx_sad32x16_c;
|
||||
if (flags & HAS_SSE2) vpx_sad32x16 = vpx_sad32x16_sse2;
|
||||
if (flags & HAS_AVX2) vpx_sad32x16 = vpx_sad32x16_avx2;
|
||||
vpx_sad32x16_avg = vpx_sad32x16_avg_c;
|
||||
if (flags & HAS_SSE2) vpx_sad32x16_avg = vpx_sad32x16_avg_sse2;
|
||||
if (flags & HAS_AVX2) vpx_sad32x16_avg = vpx_sad32x16_avg_avx2;
|
||||
vpx_sad32x16x4d = vpx_sad32x16x4d_c;
|
||||
if (flags & HAS_SSE2) vpx_sad32x16x4d = vpx_sad32x16x4d_sse2;
|
||||
vpx_sad32x32 = vpx_sad32x32_c;
|
||||
if (flags & HAS_SSE2) vpx_sad32x32 = vpx_sad32x32_sse2;
|
||||
if (flags & HAS_AVX2) vpx_sad32x32 = vpx_sad32x32_avx2;
|
||||
vpx_sad32x32_avg = vpx_sad32x32_avg_c;
|
||||
if (flags & HAS_SSE2) vpx_sad32x32_avg = vpx_sad32x32_avg_sse2;
|
||||
if (flags & HAS_AVX2) vpx_sad32x32_avg = vpx_sad32x32_avg_avx2;
|
||||
vpx_sad32x32x4d = vpx_sad32x32x4d_c;
|
||||
if (flags & HAS_SSE2) vpx_sad32x32x4d = vpx_sad32x32x4d_sse2;
|
||||
if (flags & HAS_AVX2) vpx_sad32x32x4d = vpx_sad32x32x4d_avx2;
|
||||
vpx_sad32x64 = vpx_sad32x64_c;
|
||||
if (flags & HAS_SSE2) vpx_sad32x64 = vpx_sad32x64_sse2;
|
||||
if (flags & HAS_AVX2) vpx_sad32x64 = vpx_sad32x64_avx2;
|
||||
vpx_sad32x64_avg = vpx_sad32x64_avg_c;
|
||||
if (flags & HAS_SSE2) vpx_sad32x64_avg = vpx_sad32x64_avg_sse2;
|
||||
if (flags & HAS_AVX2) vpx_sad32x64_avg = vpx_sad32x64_avg_avx2;
|
||||
vpx_sad32x64x4d = vpx_sad32x64x4d_c;
|
||||
if (flags & HAS_SSE2) vpx_sad32x64x4d = vpx_sad32x64x4d_sse2;
|
||||
vpx_sad4x4 = vpx_sad4x4_c;
|
||||
@@ -427,16 +455,21 @@ static void setup_rtcd_internal(void)
|
||||
if (flags & HAS_SSE) vpx_sad4x8x4d = vpx_sad4x8x4d_sse;
|
||||
vpx_sad64x32 = vpx_sad64x32_c;
|
||||
if (flags & HAS_SSE2) vpx_sad64x32 = vpx_sad64x32_sse2;
|
||||
if (flags & HAS_AVX2) vpx_sad64x32 = vpx_sad64x32_avx2;
|
||||
vpx_sad64x32_avg = vpx_sad64x32_avg_c;
|
||||
if (flags & HAS_SSE2) vpx_sad64x32_avg = vpx_sad64x32_avg_sse2;
|
||||
if (flags & HAS_AVX2) vpx_sad64x32_avg = vpx_sad64x32_avg_avx2;
|
||||
vpx_sad64x32x4d = vpx_sad64x32x4d_c;
|
||||
if (flags & HAS_SSE2) vpx_sad64x32x4d = vpx_sad64x32x4d_sse2;
|
||||
vpx_sad64x64 = vpx_sad64x64_c;
|
||||
if (flags & HAS_SSE2) vpx_sad64x64 = vpx_sad64x64_sse2;
|
||||
if (flags & HAS_AVX2) vpx_sad64x64 = vpx_sad64x64_avx2;
|
||||
vpx_sad64x64_avg = vpx_sad64x64_avg_c;
|
||||
if (flags & HAS_SSE2) vpx_sad64x64_avg = vpx_sad64x64_avg_sse2;
|
||||
if (flags & HAS_AVX2) vpx_sad64x64_avg = vpx_sad64x64_avg_avx2;
|
||||
vpx_sad64x64x4d = vpx_sad64x64x4d_c;
|
||||
if (flags & HAS_SSE2) vpx_sad64x64x4d = vpx_sad64x64x4d_sse2;
|
||||
if (flags & HAS_AVX2) vpx_sad64x64x4d = vpx_sad64x64x4d_avx2;
|
||||
vpx_sad8x16 = vpx_sad8x16_c;
|
||||
if (flags & HAS_MMX) vpx_sad8x16 = vpx_sad8x16_mmx;
|
||||
if (flags & HAS_SSE2) vpx_sad8x16 = vpx_sad8x16_sse2;
|
||||
@@ -468,6 +501,7 @@ static void setup_rtcd_internal(void)
|
||||
vpx_variance16x16 = vpx_variance16x16_c;
|
||||
if (flags & HAS_MMX) vpx_variance16x16 = vpx_variance16x16_mmx;
|
||||
if (flags & HAS_SSE2) vpx_variance16x16 = vpx_variance16x16_sse2;
|
||||
if (flags & HAS_AVX2) vpx_variance16x16 = vpx_variance16x16_avx2;
|
||||
vpx_variance16x32 = vpx_variance16x32_c;
|
||||
if (flags & HAS_SSE2) vpx_variance16x32 = vpx_variance16x32_sse2;
|
||||
vpx_variance16x8 = vpx_variance16x8_c;
|
||||
@@ -475,8 +509,10 @@ static void setup_rtcd_internal(void)
|
||||
if (flags & HAS_SSE2) vpx_variance16x8 = vpx_variance16x8_sse2;
|
||||
vpx_variance32x16 = vpx_variance32x16_c;
|
||||
if (flags & HAS_SSE2) vpx_variance32x16 = vpx_variance32x16_sse2;
|
||||
if (flags & HAS_AVX2) vpx_variance32x16 = vpx_variance32x16_avx2;
|
||||
vpx_variance32x32 = vpx_variance32x32_c;
|
||||
if (flags & HAS_SSE2) vpx_variance32x32 = vpx_variance32x32_sse2;
|
||||
if (flags & HAS_AVX2) vpx_variance32x32 = vpx_variance32x32_avx2;
|
||||
vpx_variance32x64 = vpx_variance32x64_c;
|
||||
if (flags & HAS_SSE2) vpx_variance32x64 = vpx_variance32x64_sse2;
|
||||
vpx_variance4x4 = vpx_variance4x4_c;
|
||||
@@ -486,8 +522,10 @@ static void setup_rtcd_internal(void)
|
||||
if (flags & HAS_SSE2) vpx_variance4x8 = vpx_variance4x8_sse2;
|
||||
vpx_variance64x32 = vpx_variance64x32_c;
|
||||
if (flags & HAS_SSE2) vpx_variance64x32 = vpx_variance64x32_sse2;
|
||||
if (flags & HAS_AVX2) vpx_variance64x32 = vpx_variance64x32_avx2;
|
||||
vpx_variance64x64 = vpx_variance64x64_c;
|
||||
if (flags & HAS_SSE2) vpx_variance64x64 = vpx_variance64x64_sse2;
|
||||
if (flags & HAS_AVX2) vpx_variance64x64 = vpx_variance64x64_avx2;
|
||||
vpx_variance8x16 = vpx_variance8x16_c;
|
||||
if (flags & HAS_MMX) vpx_variance8x16 = vpx_variance8x16_mmx;
|
||||
if (flags & HAS_SSE2) vpx_variance8x16 = vpx_variance8x16_sse2;
|
||||
|
||||
@@ -23,7 +23,8 @@ void vpx_comp_avg_pred_c(uint8_t *comp_pred, const uint8_t *pred, int width, int
|
||||
|
||||
void vpx_get16x16var_c(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, int *sum);
|
||||
void vpx_get16x16var_sse2(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, int *sum);
|
||||
#define vpx_get16x16var vpx_get16x16var_sse2
|
||||
void vpx_get16x16var_avx2(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, int *sum);
|
||||
RTCD_EXTERN void (*vpx_get16x16var)(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, int *sum);
|
||||
|
||||
unsigned int vpx_get4x4sse_cs_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride);
|
||||
#define vpx_get4x4sse_cs vpx_get4x4sse_cs_c
|
||||
@@ -41,7 +42,8 @@ unsigned int vpx_get_mb_ss_sse2(const int16_t *);
|
||||
unsigned int vpx_mse16x16_c(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int recon_stride, unsigned int *sse);
|
||||
unsigned int vpx_mse16x16_mmx(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int recon_stride, unsigned int *sse);
|
||||
unsigned int vpx_mse16x16_sse2(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int recon_stride, unsigned int *sse);
|
||||
#define vpx_mse16x16 vpx_mse16x16_sse2
|
||||
unsigned int vpx_mse16x16_avx2(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int recon_stride, unsigned int *sse);
|
||||
RTCD_EXTERN unsigned int (*vpx_mse16x16)(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int recon_stride, unsigned int *sse);
|
||||
|
||||
unsigned int vpx_mse16x8_c(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int recon_stride, unsigned int *sse);
|
||||
unsigned int vpx_mse16x8_sse2(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int recon_stride, unsigned int *sse);
|
||||
@@ -113,11 +115,13 @@ RTCD_EXTERN void (*vpx_sad16x8x8)(const uint8_t *src_ptr, int src_stride, const
|
||||
|
||||
unsigned int vpx_sad32x16_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
unsigned int vpx_sad32x16_sse2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
#define vpx_sad32x16 vpx_sad32x16_sse2
|
||||
unsigned int vpx_sad32x16_avx2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
RTCD_EXTERN unsigned int (*vpx_sad32x16)(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
|
||||
unsigned int vpx_sad32x16_avg_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
unsigned int vpx_sad32x16_avg_sse2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
#define vpx_sad32x16_avg vpx_sad32x16_avg_sse2
|
||||
unsigned int vpx_sad32x16_avg_avx2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
RTCD_EXTERN unsigned int (*vpx_sad32x16_avg)(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
|
||||
void vpx_sad32x16x4d_c(const uint8_t *src_ptr, int src_stride, const uint8_t * const ref_ptr[], int ref_stride, uint32_t *sad_array);
|
||||
void vpx_sad32x16x4d_sse2(const uint8_t *src_ptr, int src_stride, const uint8_t * const ref_ptr[], int ref_stride, uint32_t *sad_array);
|
||||
@@ -125,29 +129,34 @@ void vpx_sad32x16x4d_sse2(const uint8_t *src_ptr, int src_stride, const uint8_t
|
||||
|
||||
unsigned int vpx_sad32x32_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
unsigned int vpx_sad32x32_sse2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
#define vpx_sad32x32 vpx_sad32x32_sse2
|
||||
unsigned int vpx_sad32x32_avx2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
RTCD_EXTERN unsigned int (*vpx_sad32x32)(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
|
||||
unsigned int vpx_sad32x32_avg_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
unsigned int vpx_sad32x32_avg_sse2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
#define vpx_sad32x32_avg vpx_sad32x32_avg_sse2
|
||||
unsigned int vpx_sad32x32_avg_avx2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
RTCD_EXTERN unsigned int (*vpx_sad32x32_avg)(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
|
||||
void vpx_sad32x32x3_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, uint32_t *sad_array);
|
||||
#define vpx_sad32x32x3 vpx_sad32x32x3_c
|
||||
|
||||
void vpx_sad32x32x4d_c(const uint8_t *src_ptr, int src_stride, const uint8_t * const ref_ptr[], int ref_stride, uint32_t *sad_array);
|
||||
void vpx_sad32x32x4d_sse2(const uint8_t *src_ptr, int src_stride, const uint8_t * const ref_ptr[], int ref_stride, uint32_t *sad_array);
|
||||
#define vpx_sad32x32x4d vpx_sad32x32x4d_sse2
|
||||
void vpx_sad32x32x4d_avx2(const uint8_t *src_ptr, int src_stride, const uint8_t * const ref_ptr[], int ref_stride, uint32_t *sad_array);
|
||||
RTCD_EXTERN void (*vpx_sad32x32x4d)(const uint8_t *src_ptr, int src_stride, const uint8_t * const ref_ptr[], int ref_stride, uint32_t *sad_array);
|
||||
|
||||
void vpx_sad32x32x8_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, uint32_t *sad_array);
|
||||
#define vpx_sad32x32x8 vpx_sad32x32x8_c
|
||||
|
||||
unsigned int vpx_sad32x64_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
unsigned int vpx_sad32x64_sse2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
#define vpx_sad32x64 vpx_sad32x64_sse2
|
||||
unsigned int vpx_sad32x64_avx2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
RTCD_EXTERN unsigned int (*vpx_sad32x64)(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
|
||||
unsigned int vpx_sad32x64_avg_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
unsigned int vpx_sad32x64_avg_sse2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
#define vpx_sad32x64_avg vpx_sad32x64_avg_sse2
|
||||
unsigned int vpx_sad32x64_avg_avx2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
RTCD_EXTERN unsigned int (*vpx_sad32x64_avg)(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
|
||||
void vpx_sad32x64x4d_c(const uint8_t *src_ptr, int src_stride, const uint8_t * const ref_ptr[], int ref_stride, uint32_t *sad_array);
|
||||
void vpx_sad32x64x4d_sse2(const uint8_t *src_ptr, int src_stride, const uint8_t * const ref_ptr[], int ref_stride, uint32_t *sad_array);
|
||||
@@ -191,11 +200,13 @@ void vpx_sad4x8x8_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_p
|
||||
|
||||
unsigned int vpx_sad64x32_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
unsigned int vpx_sad64x32_sse2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
#define vpx_sad64x32 vpx_sad64x32_sse2
|
||||
unsigned int vpx_sad64x32_avx2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
RTCD_EXTERN unsigned int (*vpx_sad64x32)(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
|
||||
unsigned int vpx_sad64x32_avg_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
unsigned int vpx_sad64x32_avg_sse2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
#define vpx_sad64x32_avg vpx_sad64x32_avg_sse2
|
||||
unsigned int vpx_sad64x32_avg_avx2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
RTCD_EXTERN unsigned int (*vpx_sad64x32_avg)(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
|
||||
void vpx_sad64x32x4d_c(const uint8_t *src_ptr, int src_stride, const uint8_t * const ref_ptr[], int ref_stride, uint32_t *sad_array);
|
||||
void vpx_sad64x32x4d_sse2(const uint8_t *src_ptr, int src_stride, const uint8_t * const ref_ptr[], int ref_stride, uint32_t *sad_array);
|
||||
@@ -203,18 +214,21 @@ void vpx_sad64x32x4d_sse2(const uint8_t *src_ptr, int src_stride, const uint8_t
|
||||
|
||||
unsigned int vpx_sad64x64_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
unsigned int vpx_sad64x64_sse2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
#define vpx_sad64x64 vpx_sad64x64_sse2
|
||||
unsigned int vpx_sad64x64_avx2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
RTCD_EXTERN unsigned int (*vpx_sad64x64)(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
|
||||
unsigned int vpx_sad64x64_avg_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
unsigned int vpx_sad64x64_avg_sse2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
#define vpx_sad64x64_avg vpx_sad64x64_avg_sse2
|
||||
unsigned int vpx_sad64x64_avg_avx2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
RTCD_EXTERN unsigned int (*vpx_sad64x64_avg)(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
|
||||
void vpx_sad64x64x3_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, uint32_t *sad_array);
|
||||
#define vpx_sad64x64x3 vpx_sad64x64x3_c
|
||||
|
||||
void vpx_sad64x64x4d_c(const uint8_t *src_ptr, int src_stride, const uint8_t * const ref_ptr[], int ref_stride, uint32_t *sad_array);
|
||||
void vpx_sad64x64x4d_sse2(const uint8_t *src_ptr, int src_stride, const uint8_t * const ref_ptr[], int ref_stride, uint32_t *sad_array);
|
||||
#define vpx_sad64x64x4d vpx_sad64x64x4d_sse2
|
||||
void vpx_sad64x64x4d_avx2(const uint8_t *src_ptr, int src_stride, const uint8_t * const ref_ptr[], int ref_stride, uint32_t *sad_array);
|
||||
RTCD_EXTERN void (*vpx_sad64x64x4d)(const uint8_t *src_ptr, int src_stride, const uint8_t * const ref_ptr[], int ref_stride, uint32_t *sad_array);
|
||||
|
||||
void vpx_sad64x64x8_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, uint32_t *sad_array);
|
||||
#define vpx_sad64x64x8 vpx_sad64x64x8_c
|
||||
@@ -279,7 +293,8 @@ RTCD_EXTERN void (*vpx_sad8x8x8)(const uint8_t *src_ptr, int src_stride, const u
|
||||
unsigned int vpx_variance16x16_c(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
unsigned int vpx_variance16x16_mmx(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
unsigned int vpx_variance16x16_sse2(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
#define vpx_variance16x16 vpx_variance16x16_sse2
|
||||
unsigned int vpx_variance16x16_avx2(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
RTCD_EXTERN unsigned int (*vpx_variance16x16)(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
|
||||
unsigned int vpx_variance16x32_c(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
unsigned int vpx_variance16x32_sse2(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
@@ -292,11 +307,13 @@ unsigned int vpx_variance16x8_sse2(const uint8_t *src_ptr, int source_stride, co
|
||||
|
||||
unsigned int vpx_variance32x16_c(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
unsigned int vpx_variance32x16_sse2(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
#define vpx_variance32x16 vpx_variance32x16_sse2
|
||||
unsigned int vpx_variance32x16_avx2(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
RTCD_EXTERN unsigned int (*vpx_variance32x16)(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
|
||||
unsigned int vpx_variance32x32_c(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
unsigned int vpx_variance32x32_sse2(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
#define vpx_variance32x32 vpx_variance32x32_sse2
|
||||
unsigned int vpx_variance32x32_avx2(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
RTCD_EXTERN unsigned int (*vpx_variance32x32)(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
|
||||
unsigned int vpx_variance32x64_c(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
unsigned int vpx_variance32x64_sse2(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
@@ -313,11 +330,13 @@ unsigned int vpx_variance4x8_sse2(const uint8_t *src_ptr, int source_stride, con
|
||||
|
||||
unsigned int vpx_variance64x32_c(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
unsigned int vpx_variance64x32_sse2(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
#define vpx_variance64x32 vpx_variance64x32_sse2
|
||||
unsigned int vpx_variance64x32_avx2(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
RTCD_EXTERN unsigned int (*vpx_variance64x32)(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
|
||||
unsigned int vpx_variance64x64_c(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
unsigned int vpx_variance64x64_sse2(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
#define vpx_variance64x64 vpx_variance64x64_sse2
|
||||
unsigned int vpx_variance64x64_avx2(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
RTCD_EXTERN unsigned int (*vpx_variance64x64)(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
|
||||
unsigned int vpx_variance8x16_c(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
unsigned int vpx_variance8x16_mmx(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
@@ -343,6 +362,10 @@ static void setup_rtcd_internal(void)
|
||||
|
||||
(void)flags;
|
||||
|
||||
vpx_get16x16var = vpx_get16x16var_sse2;
|
||||
if (flags & HAS_AVX2) vpx_get16x16var = vpx_get16x16var_avx2;
|
||||
vpx_mse16x16 = vpx_mse16x16_sse2;
|
||||
if (flags & HAS_AVX2) vpx_mse16x16 = vpx_mse16x16_avx2;
|
||||
vpx_sad16x16x3 = vpx_sad16x16x3_c;
|
||||
if (flags & HAS_SSE3) vpx_sad16x16x3 = vpx_sad16x16x3_sse3;
|
||||
if (flags & HAS_SSSE3) vpx_sad16x16x3 = vpx_sad16x16x3_ssse3;
|
||||
@@ -353,10 +376,34 @@ static void setup_rtcd_internal(void)
|
||||
if (flags & HAS_SSSE3) vpx_sad16x8x3 = vpx_sad16x8x3_ssse3;
|
||||
vpx_sad16x8x8 = vpx_sad16x8x8_c;
|
||||
if (flags & HAS_SSE4_1) vpx_sad16x8x8 = vpx_sad16x8x8_sse4_1;
|
||||
vpx_sad32x16 = vpx_sad32x16_sse2;
|
||||
if (flags & HAS_AVX2) vpx_sad32x16 = vpx_sad32x16_avx2;
|
||||
vpx_sad32x16_avg = vpx_sad32x16_avg_sse2;
|
||||
if (flags & HAS_AVX2) vpx_sad32x16_avg = vpx_sad32x16_avg_avx2;
|
||||
vpx_sad32x32 = vpx_sad32x32_sse2;
|
||||
if (flags & HAS_AVX2) vpx_sad32x32 = vpx_sad32x32_avx2;
|
||||
vpx_sad32x32_avg = vpx_sad32x32_avg_sse2;
|
||||
if (flags & HAS_AVX2) vpx_sad32x32_avg = vpx_sad32x32_avg_avx2;
|
||||
vpx_sad32x32x4d = vpx_sad32x32x4d_sse2;
|
||||
if (flags & HAS_AVX2) vpx_sad32x32x4d = vpx_sad32x32x4d_avx2;
|
||||
vpx_sad32x64 = vpx_sad32x64_sse2;
|
||||
if (flags & HAS_AVX2) vpx_sad32x64 = vpx_sad32x64_avx2;
|
||||
vpx_sad32x64_avg = vpx_sad32x64_avg_sse2;
|
||||
if (flags & HAS_AVX2) vpx_sad32x64_avg = vpx_sad32x64_avg_avx2;
|
||||
vpx_sad4x4x3 = vpx_sad4x4x3_c;
|
||||
if (flags & HAS_SSE3) vpx_sad4x4x3 = vpx_sad4x4x3_sse3;
|
||||
vpx_sad4x4x8 = vpx_sad4x4x8_c;
|
||||
if (flags & HAS_SSE4_1) vpx_sad4x4x8 = vpx_sad4x4x8_sse4_1;
|
||||
vpx_sad64x32 = vpx_sad64x32_sse2;
|
||||
if (flags & HAS_AVX2) vpx_sad64x32 = vpx_sad64x32_avx2;
|
||||
vpx_sad64x32_avg = vpx_sad64x32_avg_sse2;
|
||||
if (flags & HAS_AVX2) vpx_sad64x32_avg = vpx_sad64x32_avg_avx2;
|
||||
vpx_sad64x64 = vpx_sad64x64_sse2;
|
||||
if (flags & HAS_AVX2) vpx_sad64x64 = vpx_sad64x64_avx2;
|
||||
vpx_sad64x64_avg = vpx_sad64x64_avg_sse2;
|
||||
if (flags & HAS_AVX2) vpx_sad64x64_avg = vpx_sad64x64_avg_avx2;
|
||||
vpx_sad64x64x4d = vpx_sad64x64x4d_sse2;
|
||||
if (flags & HAS_AVX2) vpx_sad64x64x4d = vpx_sad64x64x4d_avx2;
|
||||
vpx_sad8x16x3 = vpx_sad8x16x3_c;
|
||||
if (flags & HAS_SSE3) vpx_sad8x16x3 = vpx_sad8x16x3_sse3;
|
||||
vpx_sad8x16x8 = vpx_sad8x16x8_c;
|
||||
@@ -365,6 +412,16 @@ static void setup_rtcd_internal(void)
|
||||
if (flags & HAS_SSE3) vpx_sad8x8x3 = vpx_sad8x8x3_sse3;
|
||||
vpx_sad8x8x8 = vpx_sad8x8x8_c;
|
||||
if (flags & HAS_SSE4_1) vpx_sad8x8x8 = vpx_sad8x8x8_sse4_1;
|
||||
vpx_variance16x16 = vpx_variance16x16_sse2;
|
||||
if (flags & HAS_AVX2) vpx_variance16x16 = vpx_variance16x16_avx2;
|
||||
vpx_variance32x16 = vpx_variance32x16_sse2;
|
||||
if (flags & HAS_AVX2) vpx_variance32x16 = vpx_variance32x16_avx2;
|
||||
vpx_variance32x32 = vpx_variance32x32_sse2;
|
||||
if (flags & HAS_AVX2) vpx_variance32x32 = vpx_variance32x32_avx2;
|
||||
vpx_variance64x32 = vpx_variance64x32_sse2;
|
||||
if (flags & HAS_AVX2) vpx_variance64x32 = vpx_variance64x32_avx2;
|
||||
vpx_variance64x64 = vpx_variance64x64_sse2;
|
||||
if (flags & HAS_AVX2) vpx_variance64x64 = vpx_variance64x64_avx2;
|
||||
}
|
||||
#endif
|
||||
|
||||
|
||||
@@ -23,6 +23,7 @@ void vpx_comp_avg_pred_c(uint8_t *comp_pred, const uint8_t *pred, int width, int
|
||||
|
||||
void vpx_get16x16var_c(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, int *sum);
|
||||
void vpx_get16x16var_sse2(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, int *sum);
|
||||
void vpx_get16x16var_avx2(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, int *sum);
|
||||
RTCD_EXTERN void (*vpx_get16x16var)(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse, int *sum);
|
||||
|
||||
unsigned int vpx_get4x4sse_cs_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride);
|
||||
@@ -41,6 +42,7 @@ unsigned int vpx_get_mb_ss_sse2(const int16_t *);
|
||||
unsigned int vpx_mse16x16_c(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int recon_stride, unsigned int *sse);
|
||||
unsigned int vpx_mse16x16_mmx(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int recon_stride, unsigned int *sse);
|
||||
unsigned int vpx_mse16x16_sse2(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int recon_stride, unsigned int *sse);
|
||||
unsigned int vpx_mse16x16_avx2(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int recon_stride, unsigned int *sse);
|
||||
RTCD_EXTERN unsigned int (*vpx_mse16x16)(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int recon_stride, unsigned int *sse);
|
||||
|
||||
unsigned int vpx_mse16x8_c(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int recon_stride, unsigned int *sse);
|
||||
@@ -113,10 +115,12 @@ RTCD_EXTERN void (*vpx_sad16x8x8)(const uint8_t *src_ptr, int src_stride, const
|
||||
|
||||
unsigned int vpx_sad32x16_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
unsigned int vpx_sad32x16_sse2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
unsigned int vpx_sad32x16_avx2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
RTCD_EXTERN unsigned int (*vpx_sad32x16)(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
|
||||
unsigned int vpx_sad32x16_avg_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
unsigned int vpx_sad32x16_avg_sse2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
unsigned int vpx_sad32x16_avg_avx2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
RTCD_EXTERN unsigned int (*vpx_sad32x16_avg)(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
|
||||
void vpx_sad32x16x4d_c(const uint8_t *src_ptr, int src_stride, const uint8_t * const ref_ptr[], int ref_stride, uint32_t *sad_array);
|
||||
@@ -125,10 +129,12 @@ void vpx_sad32x16x4d_sse2(const uint8_t *src_ptr, int src_stride, const uint8_t
|
||||
|
||||
unsigned int vpx_sad32x32_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
unsigned int vpx_sad32x32_sse2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
unsigned int vpx_sad32x32_avx2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
RTCD_EXTERN unsigned int (*vpx_sad32x32)(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
|
||||
unsigned int vpx_sad32x32_avg_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
unsigned int vpx_sad32x32_avg_sse2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
unsigned int vpx_sad32x32_avg_avx2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
RTCD_EXTERN unsigned int (*vpx_sad32x32_avg)(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
|
||||
void vpx_sad32x32x3_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, uint32_t *sad_array);
|
||||
@@ -136,6 +142,7 @@ void vpx_sad32x32x3_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref
|
||||
|
||||
void vpx_sad32x32x4d_c(const uint8_t *src_ptr, int src_stride, const uint8_t * const ref_ptr[], int ref_stride, uint32_t *sad_array);
|
||||
void vpx_sad32x32x4d_sse2(const uint8_t *src_ptr, int src_stride, const uint8_t * const ref_ptr[], int ref_stride, uint32_t *sad_array);
|
||||
void vpx_sad32x32x4d_avx2(const uint8_t *src_ptr, int src_stride, const uint8_t * const ref_ptr[], int ref_stride, uint32_t *sad_array);
|
||||
RTCD_EXTERN void (*vpx_sad32x32x4d)(const uint8_t *src_ptr, int src_stride, const uint8_t * const ref_ptr[], int ref_stride, uint32_t *sad_array);
|
||||
|
||||
void vpx_sad32x32x8_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, uint32_t *sad_array);
|
||||
@@ -143,10 +150,12 @@ void vpx_sad32x32x8_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref
|
||||
|
||||
unsigned int vpx_sad32x64_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
unsigned int vpx_sad32x64_sse2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
unsigned int vpx_sad32x64_avx2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
RTCD_EXTERN unsigned int (*vpx_sad32x64)(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
|
||||
unsigned int vpx_sad32x64_avg_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
unsigned int vpx_sad32x64_avg_sse2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
unsigned int vpx_sad32x64_avg_avx2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
RTCD_EXTERN unsigned int (*vpx_sad32x64_avg)(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
|
||||
void vpx_sad32x64x4d_c(const uint8_t *src_ptr, int src_stride, const uint8_t * const ref_ptr[], int ref_stride, uint32_t *sad_array);
|
||||
@@ -191,10 +200,12 @@ void vpx_sad4x8x8_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_p
|
||||
|
||||
unsigned int vpx_sad64x32_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
unsigned int vpx_sad64x32_sse2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
unsigned int vpx_sad64x32_avx2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
RTCD_EXTERN unsigned int (*vpx_sad64x32)(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
|
||||
unsigned int vpx_sad64x32_avg_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
unsigned int vpx_sad64x32_avg_sse2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
unsigned int vpx_sad64x32_avg_avx2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
RTCD_EXTERN unsigned int (*vpx_sad64x32_avg)(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
|
||||
void vpx_sad64x32x4d_c(const uint8_t *src_ptr, int src_stride, const uint8_t * const ref_ptr[], int ref_stride, uint32_t *sad_array);
|
||||
@@ -203,10 +214,12 @@ void vpx_sad64x32x4d_sse2(const uint8_t *src_ptr, int src_stride, const uint8_t
|
||||
|
||||
unsigned int vpx_sad64x64_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
unsigned int vpx_sad64x64_sse2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
unsigned int vpx_sad64x64_avx2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
RTCD_EXTERN unsigned int (*vpx_sad64x64)(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride);
|
||||
|
||||
unsigned int vpx_sad64x64_avg_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
unsigned int vpx_sad64x64_avg_sse2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
unsigned int vpx_sad64x64_avg_avx2(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
RTCD_EXTERN unsigned int (*vpx_sad64x64_avg)(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred);
|
||||
|
||||
void vpx_sad64x64x3_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, uint32_t *sad_array);
|
||||
@@ -214,6 +227,7 @@ void vpx_sad64x64x3_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref
|
||||
|
||||
void vpx_sad64x64x4d_c(const uint8_t *src_ptr, int src_stride, const uint8_t * const ref_ptr[], int ref_stride, uint32_t *sad_array);
|
||||
void vpx_sad64x64x4d_sse2(const uint8_t *src_ptr, int src_stride, const uint8_t * const ref_ptr[], int ref_stride, uint32_t *sad_array);
|
||||
void vpx_sad64x64x4d_avx2(const uint8_t *src_ptr, int src_stride, const uint8_t * const ref_ptr[], int ref_stride, uint32_t *sad_array);
|
||||
RTCD_EXTERN void (*vpx_sad64x64x4d)(const uint8_t *src_ptr, int src_stride, const uint8_t * const ref_ptr[], int ref_stride, uint32_t *sad_array);
|
||||
|
||||
void vpx_sad64x64x8_c(const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, uint32_t *sad_array);
|
||||
@@ -279,6 +293,7 @@ RTCD_EXTERN void (*vpx_sad8x8x8)(const uint8_t *src_ptr, int src_stride, const u
|
||||
unsigned int vpx_variance16x16_c(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
unsigned int vpx_variance16x16_mmx(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
unsigned int vpx_variance16x16_sse2(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
unsigned int vpx_variance16x16_avx2(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
RTCD_EXTERN unsigned int (*vpx_variance16x16)(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
|
||||
unsigned int vpx_variance16x32_c(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
@@ -292,10 +307,12 @@ unsigned int vpx_variance16x8_sse2(const uint8_t *src_ptr, int source_stride, co
|
||||
|
||||
unsigned int vpx_variance32x16_c(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
unsigned int vpx_variance32x16_sse2(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
unsigned int vpx_variance32x16_avx2(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
RTCD_EXTERN unsigned int (*vpx_variance32x16)(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
|
||||
unsigned int vpx_variance32x32_c(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
unsigned int vpx_variance32x32_sse2(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
unsigned int vpx_variance32x32_avx2(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
RTCD_EXTERN unsigned int (*vpx_variance32x32)(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
|
||||
unsigned int vpx_variance32x64_c(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
@@ -313,10 +330,12 @@ unsigned int vpx_variance4x8_sse2(const uint8_t *src_ptr, int source_stride, con
|
||||
|
||||
unsigned int vpx_variance64x32_c(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
unsigned int vpx_variance64x32_sse2(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
unsigned int vpx_variance64x32_avx2(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
RTCD_EXTERN unsigned int (*vpx_variance64x32)(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
|
||||
unsigned int vpx_variance64x64_c(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
unsigned int vpx_variance64x64_sse2(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
unsigned int vpx_variance64x64_avx2(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
RTCD_EXTERN unsigned int (*vpx_variance64x64)(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
|
||||
unsigned int vpx_variance8x16_c(const uint8_t *src_ptr, int source_stride, const uint8_t *ref_ptr, int ref_stride, unsigned int *sse);
|
||||
@@ -344,7 +363,9 @@ static void setup_rtcd_internal(void)
|
||||
(void)flags;
|
||||
|
||||
vpx_get16x16var = vpx_get16x16var_sse2;
|
||||
if (flags & HAS_AVX2) vpx_get16x16var = vpx_get16x16var_avx2;
|
||||
vpx_mse16x16 = vpx_mse16x16_sse2;
|
||||
if (flags & HAS_AVX2) vpx_mse16x16 = vpx_mse16x16_avx2;
|
||||
vpx_sad16x16x3 = vpx_sad16x16x3_c;
|
||||
if (flags & HAS_SSE3) vpx_sad16x16x3 = vpx_sad16x16x3_sse3;
|
||||
if (flags & HAS_SSSE3) vpx_sad16x16x3 = vpx_sad16x16x3_ssse3;
|
||||
@@ -356,21 +377,33 @@ static void setup_rtcd_internal(void)
|
||||
vpx_sad16x8x8 = vpx_sad16x8x8_c;
|
||||
if (flags & HAS_SSE4_1) vpx_sad16x8x8 = vpx_sad16x8x8_sse4_1;
|
||||
vpx_sad32x16 = vpx_sad32x16_sse2;
|
||||
if (flags & HAS_AVX2) vpx_sad32x16 = vpx_sad32x16_avx2;
|
||||
vpx_sad32x16_avg = vpx_sad32x16_avg_sse2;
|
||||
if (flags & HAS_AVX2) vpx_sad32x16_avg = vpx_sad32x16_avg_avx2;
|
||||
vpx_sad32x32 = vpx_sad32x32_sse2;
|
||||
if (flags & HAS_AVX2) vpx_sad32x32 = vpx_sad32x32_avx2;
|
||||
vpx_sad32x32_avg = vpx_sad32x32_avg_sse2;
|
||||
if (flags & HAS_AVX2) vpx_sad32x32_avg = vpx_sad32x32_avg_avx2;
|
||||
vpx_sad32x32x4d = vpx_sad32x32x4d_sse2;
|
||||
if (flags & HAS_AVX2) vpx_sad32x32x4d = vpx_sad32x32x4d_avx2;
|
||||
vpx_sad32x64 = vpx_sad32x64_sse2;
|
||||
if (flags & HAS_AVX2) vpx_sad32x64 = vpx_sad32x64_avx2;
|
||||
vpx_sad32x64_avg = vpx_sad32x64_avg_sse2;
|
||||
if (flags & HAS_AVX2) vpx_sad32x64_avg = vpx_sad32x64_avg_avx2;
|
||||
vpx_sad4x4x3 = vpx_sad4x4x3_c;
|
||||
if (flags & HAS_SSE3) vpx_sad4x4x3 = vpx_sad4x4x3_sse3;
|
||||
vpx_sad4x4x8 = vpx_sad4x4x8_c;
|
||||
if (flags & HAS_SSE4_1) vpx_sad4x4x8 = vpx_sad4x4x8_sse4_1;
|
||||
vpx_sad64x32 = vpx_sad64x32_sse2;
|
||||
if (flags & HAS_AVX2) vpx_sad64x32 = vpx_sad64x32_avx2;
|
||||
vpx_sad64x32_avg = vpx_sad64x32_avg_sse2;
|
||||
if (flags & HAS_AVX2) vpx_sad64x32_avg = vpx_sad64x32_avg_avx2;
|
||||
vpx_sad64x64 = vpx_sad64x64_sse2;
|
||||
if (flags & HAS_AVX2) vpx_sad64x64 = vpx_sad64x64_avx2;
|
||||
vpx_sad64x64_avg = vpx_sad64x64_avg_sse2;
|
||||
if (flags & HAS_AVX2) vpx_sad64x64_avg = vpx_sad64x64_avg_avx2;
|
||||
vpx_sad64x64x4d = vpx_sad64x64x4d_sse2;
|
||||
if (flags & HAS_AVX2) vpx_sad64x64x4d = vpx_sad64x64x4d_avx2;
|
||||
vpx_sad8x16x3 = vpx_sad8x16x3_c;
|
||||
if (flags & HAS_SSE3) vpx_sad8x16x3 = vpx_sad8x16x3_sse3;
|
||||
vpx_sad8x16x8 = vpx_sad8x16x8_c;
|
||||
@@ -380,10 +413,15 @@ static void setup_rtcd_internal(void)
|
||||
vpx_sad8x8x8 = vpx_sad8x8x8_c;
|
||||
if (flags & HAS_SSE4_1) vpx_sad8x8x8 = vpx_sad8x8x8_sse4_1;
|
||||
vpx_variance16x16 = vpx_variance16x16_sse2;
|
||||
if (flags & HAS_AVX2) vpx_variance16x16 = vpx_variance16x16_avx2;
|
||||
vpx_variance32x16 = vpx_variance32x16_sse2;
|
||||
if (flags & HAS_AVX2) vpx_variance32x16 = vpx_variance32x16_avx2;
|
||||
vpx_variance32x32 = vpx_variance32x32_sse2;
|
||||
if (flags & HAS_AVX2) vpx_variance32x32 = vpx_variance32x32_avx2;
|
||||
vpx_variance64x32 = vpx_variance64x32_sse2;
|
||||
if (flags & HAS_AVX2) vpx_variance64x32 = vpx_variance64x32_avx2;
|
||||
vpx_variance64x64 = vpx_variance64x64_sse2;
|
||||
if (flags & HAS_AVX2) vpx_variance64x64 = vpx_variance64x64_avx2;
|
||||
}
|
||||
#endif
|
||||
|
||||
|
||||
@@ -1,5 +1,5 @@
|
||||
/*
|
||||
* Copyright (c) 2011 The WebM project authors. All Rights Reserved.
|
||||
* Copyright (c) 2015 The WebM project authors. All Rights Reserved.
|
||||
*
|
||||
* Use of this source code is governed by a BSD-style license
|
||||
* that can be found in the LICENSE file in the root of the source
|
||||
@@ -13,63 +13,83 @@
|
||||
|
||||
#include "vpx_config.h"
|
||||
|
||||
/* Implement a function wrapper to guarantee initialization
|
||||
* thread-safety for library singletons.
|
||||
*
|
||||
* NOTE: These functions use static locks, and can only be
|
||||
* used with one common argument per compilation unit. So
|
||||
*
|
||||
* file1.c:
|
||||
* vpx_once(foo);
|
||||
* ...
|
||||
* vpx_once(foo);
|
||||
*
|
||||
* file2.c:
|
||||
* vpx_once(bar);
|
||||
*
|
||||
* will ensure foo() and bar() are each called only once, but in
|
||||
*
|
||||
* file1.c:
|
||||
* vpx_once(foo);
|
||||
* vpx_once(bar):
|
||||
*
|
||||
* bar() will never be called because the lock is used up
|
||||
* by the call to foo().
|
||||
*/
|
||||
|
||||
#if CONFIG_MULTITHREAD && defined(_WIN32)
|
||||
#include <windows.h>
|
||||
#include <stdlib.h>
|
||||
/* Declare a per-compilation-unit state variable to track the progress
|
||||
* of calling func() only once. This must be at global scope because
|
||||
* local initializers are not thread-safe in MSVC prior to Visual
|
||||
* Studio 2015.
|
||||
*
|
||||
* As a static, once_state will be zero-initialized as program start.
|
||||
*/
|
||||
static LONG once_state;
|
||||
static void once(void (*func)(void))
|
||||
{
|
||||
static CRITICAL_SECTION *lock;
|
||||
static LONG waiters;
|
||||
static int done;
|
||||
void *lock_ptr = &lock;
|
||||
|
||||
/* If the initialization is complete, return early. This isn't just an
|
||||
* optimization, it prevents races on the destruction of the global
|
||||
* lock.
|
||||
/* Try to advance once_state from its initial value of 0 to 1.
|
||||
* Only one thread can succeed in doing so.
|
||||
*/
|
||||
if(done)
|
||||
return;
|
||||
|
||||
InterlockedIncrement(&waiters);
|
||||
|
||||
/* Get a lock. We create one and try to make it the one-true-lock,
|
||||
* throwing it away if we lost the race.
|
||||
*/
|
||||
|
||||
{
|
||||
/* Scope to protect access to new_lock */
|
||||
CRITICAL_SECTION *new_lock = malloc(sizeof(CRITICAL_SECTION));
|
||||
InitializeCriticalSection(new_lock);
|
||||
if (InterlockedCompareExchangePointer(lock_ptr, new_lock, NULL) != NULL)
|
||||
{
|
||||
DeleteCriticalSection(new_lock);
|
||||
free(new_lock);
|
||||
}
|
||||
}
|
||||
|
||||
/* At this point, we have a lock that can be synchronized on. We don't
|
||||
* care which thread actually performed the allocation.
|
||||
*/
|
||||
|
||||
EnterCriticalSection(lock);
|
||||
|
||||
if (!done)
|
||||
{
|
||||
if (InterlockedCompareExchange(&once_state, 1, 0) == 0) {
|
||||
/* We're the winning thread, having set once_state to 1.
|
||||
* Call our function. */
|
||||
func();
|
||||
done = 1;
|
||||
/* Now advance once_state to 2, unblocking any other threads. */
|
||||
InterlockedIncrement(&once_state);
|
||||
return;
|
||||
}
|
||||
|
||||
LeaveCriticalSection(lock);
|
||||
|
||||
/* Last one out should free resources. The destructed objects are
|
||||
* protected by checking if(done) above.
|
||||
/* We weren't the winning thread, but we want to block on
|
||||
* the state variable so we don't return before func()
|
||||
* has finished executing elsewhere.
|
||||
*
|
||||
* Try to advance once_state from 2 to 2, which is only possible
|
||||
* after the winning thead advances it from 1 to 2.
|
||||
*/
|
||||
if(!InterlockedDecrement(&waiters))
|
||||
{
|
||||
DeleteCriticalSection(lock);
|
||||
free(lock);
|
||||
lock = NULL;
|
||||
while (InterlockedCompareExchange(&once_state, 2, 2) != 2) {
|
||||
/* State isn't yet 2. Try again.
|
||||
*
|
||||
* We are used for singleton initialization functions,
|
||||
* which should complete quickly. Contention will likewise
|
||||
* be rare, so it's worthwhile to use a simple but cpu-
|
||||
* intensive busy-wait instead of successive backoff,
|
||||
* waiting on a kernel object, or another heavier-weight scheme.
|
||||
*
|
||||
* We can at least yield our timeslice.
|
||||
*/
|
||||
Sleep(0);
|
||||
}
|
||||
|
||||
/* We've seen once_state advance to 2, so we know func()
|
||||
* has been called. And we've left once_state as we found it,
|
||||
* so other threads will have the same experience.
|
||||
*
|
||||
* It's safe to return now.
|
||||
*/
|
||||
return;
|
||||
}
|
||||
|
||||
|
||||
|
||||
@@ -11,6 +11,9 @@
|
||||
#define RTCD_C
|
||||
#include "./vpx_scale_rtcd.h"
|
||||
#include "vpx_ports/vpx_once.h"
|
||||
#ifdef _MSC_VER
|
||||
#include <intrin.h>
|
||||
#endif
|
||||
|
||||
void vpx_scale_rtcd()
|
||||
{
|
||||
|
||||
Reference in New Issue
Block a user