mozilla
/
gecko-dev
зеркало из https://github.com/mozilla/gecko-dev.git
1
0
Форкнуть 0
Код Задачи Пакеты Проекты Релизы Вики Активность

Bug 1178215 - Update libvpx. r=rillian 

Update libvpx to e67d45d4ce92468ba193288b59093fef0a502662
* * *
Bug 1178215 - Bump clobber for libvpx file renames. r=me

--HG--
rename : media/libvpx/vp8/common/x86/variance_impl_mmx.asm => media/libvpx/vp8/common/x86/vp8_variance_impl_mmx.asm
rename : media/libvpx/vp8/common/x86/variance_mmx.c => media/libvpx/vp8/common/x86/vp8_variance_mmx.c
rename : media/libvpx/vp8/common/x86/variance_sse2.c => media/libvpx/vp8/common/x86/vp8_variance_sse2.c
rename : media/libvpx/vp9/common/arm/neon/vp9_avg_neon.c => media/libvpx/vp9/common/arm/neon/vp9_convolve_avg_neon.c
rename : media/libvpx/vp9/common/arm/neon/vp9_avg_neon_asm.asm => media/libvpx/vp9/common/arm/neon/vp9_convolve_avg_neon_asm.asm
rename : media/libvpx/vp9/encoder/x86/vp9_dct32x32_avx2.c => media/libvpx/vp9/encoder/x86/vp9_dct32x32_avx2_impl.h
rename : media/libvpx/vp9/encoder/x86/vp9_dct32x32_sse2.c => media/libvpx/vp9/encoder/x86/vp9_dct32x32_sse2_impl.h
rename : media/libvpx/vp9/encoder/x86/vp9_dct_impl_sse2.c => media/libvpx/vp9/encoder/x86/vp9_dct_sse2_impl.h
rename : media/libvpx/vp9/encoder/arm/neon/vp9_sad4d_neon.c => media/libvpx/vpx_dsp/arm/sad4d_neon.c
rename : media/libvpx/vp8/common/arm/armv6/vp8_sad16x16_armv6.asm => media/libvpx/vpx_dsp/arm/sad_media.asm
rename : media/libvpx/vp9/encoder/arm/neon/vp9_sad_neon.c => media/libvpx/vpx_dsp/arm/sad_neon.c
rename : media/libvpx/vp8/common/arm/armv6/vp8_variance16x16_armv6.asm => media/libvpx/vpx_dsp/arm/variance_media.asm
rename : media/libvpx/vp9/encoder/vp9_sad.c => media/libvpx/vpx_dsp/sad.c
rename : media/libvpx/vp9/encoder/x86/vp9_highbd_sad4d_sse2.asm => media/libvpx/vpx_dsp/x86/highbd_sad4d_sse2.asm
rename : media/libvpx/vp9/encoder/x86/vp9_highbd_sad_sse2.asm => media/libvpx/vpx_dsp/x86/highbd_sad_sse2.asm
rename : media/libvpx/vp9/encoder/x86/vp9_highbd_variance_impl_sse2.asm => media/libvpx/vpx_dsp/x86/highbd_variance_impl_sse2.asm
rename : media/libvpx/vp9/encoder/x86/vp9_sad4d_intrin_avx2.c => media/libvpx/vpx_dsp/x86/sad4d_avx2.c
rename : media/libvpx/vp9/encoder/x86/vp9_sad4d_sse2.asm => media/libvpx/vpx_dsp/x86/sad4d_sse2.asm
rename : media/libvpx/vp9/encoder/x86/vp9_sad_intrin_avx2.c => media/libvpx/vpx_dsp/x86/sad_avx2.c
rename : media/libvpx/vp8/common/x86/sad_mmx.asm => media/libvpx/vpx_dsp/x86/sad_mmx.asm
rename : media/libvpx/vp9/encoder/x86/vp9_sad_sse2.asm => media/libvpx/vpx_dsp/x86/sad_sse2.asm
rename : media/libvpx/vp9/encoder/x86/vp9_sad_sse3.asm => media/libvpx/vpx_dsp/x86/sad_sse3.asm
rename : media/libvpx/vp9/encoder/x86/vp9_sad_sse4.asm => media/libvpx/vpx_dsp/x86/sad_sse4.asm
rename : media/libvpx/vp9/encoder/x86/vp9_sad_ssse3.asm => media/libvpx/vpx_dsp/x86/sad_ssse3.asm
rename : media/libvpx/vp9/encoder/x86/vp9_variance_impl_intrin_avx2.c => media/libvpx/vpx_dsp/x86/variance_impl_avx2.c
rename : media/libvpx/vp8/common/x86/variance_impl_mmx.asm => media/libvpx/vpx_dsp/x86/variance_impl_mmx.asm
This commit is contained in:
Jan Gerber 2015-06-29 23:07:20 +02:00
Родитель c186adf8ce
Коммит 0741d70e20
310 изменённых файлов: 16687 добавлений и 18757 удалений
Показать статистику Скачать Patch файл Скачать Diff файл Показать все файлы Свернуть все файлы

2
CLOBBER
Просмотреть файл

@ -22,4 +22,4 @@
# changes to stick? As of bug 928195, this shouldn't be necessary! Please
# don't change CLOBBER for WebIDL changes any more.
Bug 1075758 (updating ICU to 55) requires a clobber.
Bug 1178215 requires clobber for libvpx file moves.

2
media/libvpx/PATENTS
Просмотреть файл

@ -17,7 +17,7 @@ or agree to the institution of patent litigation or any other patent
enforcement activity against any entity (including a cross-claim or
counterclaim in a lawsuit) alleging that any of these implementations of WebM
or any code incorporated within any of these implementations of WebM
constitutes direct or contributory patent infringement, or inducement of
constitute direct or contributory patent infringement, or inducement of
patent infringement, then any patent rights granted to you under this License
for these implementations of WebM shall terminate as of the date such
litigation is filed.

2
media/libvpx/README_MOZILLA
Просмотреть файл

@ -8,4 +8,4 @@ The libvpx git repository is:
https://chromium.googlesource.com/webm/libvpx
The git commit ID used was c74bf6d889992c3cabe017ec353ca85c323107cd
The git commit ID used was e67d45d4ce92468ba193288b59093fef0a502662

63
media/libvpx/sources.mozbuild
Просмотреть файл

@ -14,9 +14,6 @@ files = {
'vp8/common/arm/armv6/loopfilter_v6.asm',
'vp8/common/arm/armv6/simpleloopfilter_v6.asm',
'vp8/common/arm/armv6/sixtappredict8x4_v6.asm',
'vp8/common/arm/armv6/vp8_sad16x16_armv6.asm',
'vp8/common/arm/armv6/vp8_variance16x16_armv6.asm',
'vp8/common/arm/armv6/vp8_variance8x8_armv6.asm',
'vp8/common/arm/armv6/vp8_variance_halfpixvar16x16_h_armv6.asm',
'vp8/common/arm/armv6/vp8_variance_halfpixvar16x16_hv_armv6.asm',
'vp8/common/arm/armv6/vp8_variance_halfpixvar16x16_v_armv6.asm',
@ -38,13 +35,10 @@ files = {
'vp8/common/arm/neon/loopfiltersimpleverticaledge_neon.c',
'vp8/common/arm/neon/mbloopfilter_neon.c',
'vp8/common/arm/neon/reconintra_neon.c',
'vp8/common/arm/neon/sad_neon.c',
'vp8/common/arm/neon/shortidct4x4llm_neon.c',
'vp8/common/arm/neon/sixtappredict_neon.c',
'vp8/common/arm/neon/variance_neon.c',
'vp8/common/arm/neon/vp8_subpixelvariance_neon.c',
'vp8/common/arm/variance_arm.c',
'vp8/encoder/arm/armv6/vp8_mse16x16_armv6.asm',
'vp8/encoder/arm/armv6/vp8_short_fdct4x4_armv6.asm',
'vp8/encoder/arm/armv6/walsh_v6.asm',
'vp8/encoder/arm/dct_arm.c',
@ -52,11 +46,10 @@ files = {
'vp8/encoder/arm/neon/fastquantizeb_neon.c',
'vp8/encoder/arm/neon/shortfdct_neon.c',
'vp8/encoder/arm/neon/subtract_neon.c',
'vp8/encoder/arm/neon/vp8_mse16x16_neon.c',
'vp8/encoder/arm/neon/vp8_shortwalsh4x4_neon.c',
'vp9/common/arm/neon/vp9_avg_neon_asm.asm',
'vp9/common/arm/neon/vp9_convolve8_avg_neon_asm.asm',
'vp9/common/arm/neon/vp9_convolve8_neon_asm.asm',
'vp9/common/arm/neon/vp9_convolve_avg_neon_asm.asm',
'vp9/common/arm/neon/vp9_convolve_neon.c',
'vp9/common/arm/neon/vp9_copy_neon_asm.asm',
'vp9/common/arm/neon/vp9_idct16x16_1_add_neon_asm.asm',
@ -75,26 +68,30 @@ files = {
'vp9/common/arm/neon/vp9_loopfilter_8_neon_asm.asm',
'vp9/common/arm/neon/vp9_loopfilter_neon.c',
'vp9/common/arm/neon/vp9_mb_lpf_neon.asm',
'vp9/common/arm/neon/vp9_reconintra_neon.c',
'vp9/common/arm/neon/vp9_reconintra_neon_asm.asm',
'vp9/common/arm/neon/vp9_save_reg_neon.asm',
'vp9/encoder/arm/neon/vp9_dct_neon.c',
'vp9/encoder/arm/neon/vp9_quantize_neon.c',
'vp9/encoder/arm/neon/vp9_sad4d_neon.c',
'vp9/encoder/arm/neon/vp9_sad_neon.c',
'vp9/encoder/arm/neon/vp9_subtract_neon.c',
'vp9/encoder/arm/neon/vp9_variance_neon.c',
'vp9/encoder/arm/neon/vp9enc_avg_neon.c',
'vpx_dsp/arm/sad4d_neon.c',
'vpx_dsp/arm/sad_media.asm',
'vpx_dsp/arm/sad_neon.c',
'vpx_dsp/arm/variance_media.asm',
'vpx_dsp/arm/variance_neon.c',
'vpx_ports/arm_cpudetect.c'],
'AVX2': ['vp9/common/x86/vp9_loopfilter_intrin_avx2.c',
'vp9/common/x86/vp9_subpixel_8t_intrin_avx2.c',
'vp9/encoder/x86/vp9_dct32x32_avx2.c',
'vp9/encoder/x86/vp9_dct_avx2.c',
'vp9/encoder/x86/vp9_error_intrin_avx2.c',
'vp9/encoder/x86/vp9_sad4d_intrin_avx2.c',
'vp9/encoder/x86/vp9_sad_intrin_avx2.c',
'vp9/encoder/x86/vp9_subpel_variance_impl_intrin_avx2.c',
'vp9/encoder/x86/vp9_variance_avx2.c',
'vp9/encoder/x86/vp9_variance_impl_intrin_avx2.c'],
'vpx_dsp/x86/sad4d_avx2.c',
'vpx_dsp/x86/sad_avx2.c',
'vpx_dsp/x86/variance_avx2.c',
'vpx_dsp/x86/variance_impl_avx2.c'],
'ERROR_CONCEALMENT': ['vp8/decoder/error_concealment.c'],
'EXPORTS': ['vpx/vp8.h',
'vpx/vp8cx.h',
@ -115,6 +112,7 @@ files = {
'vpx_scale/yv12config.h'],
'SOURCES': ['vp8/common/alloccommon.c',
'vp8/common/blockd.c',
'vp8/common/copy_c.c',
'vp8/common/debugmodes.c',
'vp8/common/dequantize.c',
'vp8/common/entropy.c',
@ -135,7 +133,6 @@ files = {
'vp8/common/reconintra.c',
'vp8/common/reconintra4x4.c',
'vp8/common/rtcd.c',
'vp8/common/sad_c.c',
'vp8/common/setupintrarecon.c',
'vp8/common/swapyv12buffer.c',
'vp8/common/treecoder.c',
@ -230,7 +227,6 @@ files = {
'vp9/encoder/vp9_rd.c',
'vp9/encoder/vp9_rdopt.c',
'vp9/encoder/vp9_resize.c',
'vp9/encoder/vp9_sad.c',
'vp9/encoder/vp9_segmentation.c',
'vp9/encoder/vp9_skin_detection.c',
'vp9/encoder/vp9_speed_features.c',
@ -249,6 +245,9 @@ files = {
'vpx/src/vpx_encoder.c',
'vpx/src/vpx_image.c',
'vpx/src/vpx_psnr.c',
'vpx_dsp/sad.c',
'vpx_dsp/variance.c',
'vpx_dsp/vpx_dsp_rtcd.c',
'vpx_mem/vpx_mem.c',
'vpx_scale/generic/gen_scalers.c',
'vpx_scale/generic/vpx_scale.c',
@ -266,7 +265,9 @@ files = {
'vp9/encoder/x86/vp9_dct_ssse3_x86_64.asm',
'vp9/encoder/x86/vp9_quantize_ssse3_x86_64.asm',
'vp9/encoder/x86/vp9_ssim_opt_x86_64.asm'],
'X86_ASM': ['vp8/common/x86/dequantize_mmx.asm',
'X86_ASM': ['vp8/common/x86/copy_sse2.asm',
'vp8/common/x86/copy_sse3.asm',
'vp8/common/x86/dequantize_mmx.asm',
'vp8/common/x86/filter_x86.c',
'vp8/common/x86/idct_blk_mmx.c',
'vp8/common/x86/idct_blk_sse2.c',
@ -283,21 +284,16 @@ files = {
'vp8/common/x86/recon_mmx.asm',
'vp8/common/x86/recon_sse2.asm',
'vp8/common/x86/recon_wrapper_sse2.c',
'vp8/common/x86/sad_mmx.asm',
'vp8/common/x86/sad_sse2.asm',
'vp8/common/x86/sad_sse3.asm',
'vp8/common/x86/sad_sse4.asm',
'vp8/common/x86/sad_ssse3.asm',
'vp8/common/x86/subpixel_mmx.asm',
'vp8/common/x86/subpixel_sse2.asm',
'vp8/common/x86/subpixel_ssse3.asm',
'vp8/common/x86/variance_impl_mmx.asm',
'vp8/common/x86/variance_impl_sse2.asm',
'vp8/common/x86/variance_impl_ssse3.asm',
'vp8/common/x86/variance_mmx.c',
'vp8/common/x86/variance_sse2.c',
'vp8/common/x86/variance_ssse3.c',
'vp8/common/x86/vp8_asm_stubs.c',
'vp8/common/x86/vp8_variance_impl_mmx.asm',
'vp8/common/x86/vp8_variance_mmx.c',
'vp8/common/x86/vp8_variance_sse2.c',
'vp8/encoder/x86/dct_mmx.asm',
'vp8/encoder/x86/dct_sse2.asm',
'vp8/encoder/x86/denoising_sse2.c',
@ -315,7 +311,6 @@ files = {
'vp9/common/x86/vp9_asm_stubs.c',
'vp9/common/x86/vp9_copy_sse2.asm',
'vp9/common/x86/vp9_idct_intrin_sse2.c',
'vp9/common/x86/vp9_idct_intrin_ssse3.c',
'vp9/common/x86/vp9_intrapred_sse2.asm',
'vp9/common/x86/vp9_intrapred_ssse3.asm',
'vp9/common/x86/vp9_loopfilter_intrin_sse2.c',
@ -326,21 +321,23 @@ files = {
'vp9/common/x86/vp9_subpixel_bilinear_sse2.asm',
'vp9/common/x86/vp9_subpixel_bilinear_ssse3.asm',
'vp9/encoder/x86/vp9_avg_intrin_sse2.c',
'vp9/encoder/x86/vp9_dct32x32_sse2.c',
'vp9/encoder/x86/vp9_dct_impl_sse2.c',
'vp9/encoder/x86/vp9_dct_mmx.asm',
'vp9/encoder/x86/vp9_dct_sse2.c',
'vp9/encoder/x86/vp9_dct_ssse3.c',
'vp9/encoder/x86/vp9_error_sse2.asm',
'vp9/encoder/x86/vp9_quantize_sse2.c',
'vp9/encoder/x86/vp9_sad4d_sse2.asm',
'vp9/encoder/x86/vp9_sad_sse2.asm',
'vp9/encoder/x86/vp9_sad_sse3.asm',
'vp9/encoder/x86/vp9_sad_sse4.asm',
'vp9/encoder/x86/vp9_sad_ssse3.asm',
'vp9/encoder/x86/vp9_subpel_variance.asm',
'vp9/encoder/x86/vp9_subtract_sse2.asm',
'vp9/encoder/x86/vp9_temporal_filter_apply_sse2.asm',
'vp9/encoder/x86/vp9_variance_sse2.c',
'vpx_dsp/x86/sad4d_sse2.asm',
'vpx_dsp/x86/sad_mmx.asm',
'vpx_dsp/x86/sad_sse2.asm',
'vpx_dsp/x86/sad_sse3.asm',
'vpx_dsp/x86/sad_sse4.asm',
'vpx_dsp/x86/sad_ssse3.asm',
'vpx_dsp/x86/variance_impl_mmx.asm',
'vpx_dsp/x86/variance_mmx.c',
'vpx_dsp/x86/variance_sse2.c',
'vpx_ports/emms.asm']
}

2
media/libvpx/third_party/x86inc/x86inc.asm поставляемый
Просмотреть файл

@ -36,7 +36,9 @@
%include "vpx_config.asm"
%ifndef program_name
%define program_name vp9
%endif
%define UNIX64 0

9
media/libvpx/vp8/common/alloccommon.c
Просмотреть файл

@ -10,6 +10,7 @@
#include "vpx_config.h"
#include "alloccommon.h"
#include "blockd.h"
#include "vpx_mem/vpx_mem.h"
#include "onyxc_int.h"
@ -103,9 +104,9 @@ int vp8_alloc_frame_buffers(VP8_COMMON *oci, int width, int height)
goto allocation_fail;
oci->post_proc_buffer_int_used = 0;
vpx_memset(&oci->postproc_state, 0, sizeof(oci->postproc_state));
vpx_memset(oci->post_proc_buffer.buffer_alloc, 128,
oci->post_proc_buffer.frame_size);
memset(&oci->postproc_state, 0, sizeof(oci->postproc_state));
memset(oci->post_proc_buffer.buffer_alloc, 128,
oci->post_proc_buffer.frame_size);
/* Allocate buffer to store post-processing filter coefficients.
*
@ -176,7 +177,7 @@ void vp8_create_common(VP8_COMMON *oci)
oci->clamp_type = RECON_CLAMP_REQUIRED;
/* Initialize reference frame sign bias structure to defaults */
vpx_memset(oci->ref_frame_sign_bias, 0, sizeof(oci->ref_frame_sign_bias));
memset(oci->ref_frame_sign_bias, 0, sizeof(oci->ref_frame_sign_bias));
/* Default disable buffer to buffer copying */
oci->copy_buffer_to_gf = 0;

2
media/libvpx/vp8/common/arm/armv6/dequant_idct_v6.asm
Просмотреть файл

@ -165,7 +165,7 @@ vp8_dequant_idct_loop2_v6
str r1, [r2], r12 ; store output to dst
bne vp8_dequant_idct_loop2_v6
; vpx_memset
; memset
sub r0, r0, #32
add sp, sp, #4

154
media/libvpx/vp8/common/arm/armv6/vp8_variance16x16_armv6.asm
Просмотреть файл

@ -1,154 +0,0 @@
;
; Copyright (c) 2011 The WebM project authors. All Rights Reserved.
;
; Use of this source code is governed by a BSD-style license
; that can be found in the LICENSE file in the root of the source
; tree. An additional intellectual property rights grant can be found
; in the file PATENTS. All contributing project authors may
; be found in the AUTHORS file in the root of the source tree.
;
EXPORT |vp8_variance16x16_armv6|
ARM
REQUIRE8
PRESERVE8
AREA ||.text||, CODE, READONLY, ALIGN=2
; r0 unsigned char *src_ptr
; r1 int source_stride
; r2 unsigned char *ref_ptr
; r3 int recon_stride
; stack unsigned int *sse
|vp8_variance16x16_armv6| PROC
stmfd sp!, {r4-r12, lr}
pld [r0, r1, lsl #0]
pld [r2, r3, lsl #0]
mov r8, #0 ; initialize sum = 0
mov r11, #0 ; initialize sse = 0
mov r12, #16 ; set loop counter to 16 (=block height)
loop
; 1st 4 pixels
ldr r4, [r0, #0] ; load 4 src pixels
ldr r5, [r2, #0] ; load 4 ref pixels
mov lr, #0 ; constant zero
usub8 r6, r4, r5 ; calculate difference
pld [r0, r1, lsl #1]
sel r7, r6, lr ; select bytes with positive difference
usub8 r9, r5, r4 ; calculate difference with reversed operands
pld [r2, r3, lsl #1]
sel r6, r9, lr ; select bytes with negative difference
; calculate partial sums
usad8 r4, r7, lr ; calculate sum of positive differences
usad8 r5, r6, lr ; calculate sum of negative differences
orr r6, r6, r7 ; differences of all 4 pixels
; calculate total sum
adds r8, r8, r4 ; add positive differences to sum
subs r8, r8, r5 ; subtract negative differences from sum
; calculate sse
uxtb16 r5, r6 ; byte (two pixels) to halfwords
uxtb16 r10, r6, ror #8 ; another two pixels to halfwords
smlad r11, r5, r5, r11 ; dual signed multiply, add and accumulate (1)
; 2nd 4 pixels
ldr r4, [r0, #4] ; load 4 src pixels
ldr r5, [r2, #4] ; load 4 ref pixels
smlad r11, r10, r10, r11 ; dual signed multiply, add and accumulate (2)
usub8 r6, r4, r5 ; calculate difference
sel r7, r6, lr ; select bytes with positive difference
usub8 r9, r5, r4 ; calculate difference with reversed operands
sel r6, r9, lr ; select bytes with negative difference
; calculate partial sums
usad8 r4, r7, lr ; calculate sum of positive differences
usad8 r5, r6, lr ; calculate sum of negative differences
orr r6, r6, r7 ; differences of all 4 pixels
; calculate total sum
add r8, r8, r4 ; add positive differences to sum
sub r8, r8, r5 ; subtract negative differences from sum
; calculate sse
uxtb16 r5, r6 ; byte (two pixels) to halfwords
uxtb16 r10, r6, ror #8 ; another two pixels to halfwords
smlad r11, r5, r5, r11 ; dual signed multiply, add and accumulate (1)
; 3rd 4 pixels
ldr r4, [r0, #8] ; load 4 src pixels
ldr r5, [r2, #8] ; load 4 ref pixels
smlad r11, r10, r10, r11 ; dual signed multiply, add and accumulate (2)
usub8 r6, r4, r5 ; calculate difference
sel r7, r6, lr ; select bytes with positive difference
usub8 r9, r5, r4 ; calculate difference with reversed operands
sel r6, r9, lr ; select bytes with negative difference
; calculate partial sums
usad8 r4, r7, lr ; calculate sum of positive differences
usad8 r5, r6, lr ; calculate sum of negative differences
orr r6, r6, r7 ; differences of all 4 pixels
; calculate total sum
add r8, r8, r4 ; add positive differences to sum
sub r8, r8, r5 ; subtract negative differences from sum
; calculate sse
uxtb16 r5, r6 ; byte (two pixels) to halfwords
uxtb16 r10, r6, ror #8 ; another two pixels to halfwords
smlad r11, r5, r5, r11 ; dual signed multiply, add and accumulate (1)
; 4th 4 pixels
ldr r4, [r0, #12] ; load 4 src pixels
ldr r5, [r2, #12] ; load 4 ref pixels
smlad r11, r10, r10, r11 ; dual signed multiply, add and accumulate (2)
usub8 r6, r4, r5 ; calculate difference
add r0, r0, r1 ; set src_ptr to next row
sel r7, r6, lr ; select bytes with positive difference
usub8 r9, r5, r4 ; calculate difference with reversed operands
add r2, r2, r3 ; set dst_ptr to next row
sel r6, r9, lr ; select bytes with negative difference
; calculate partial sums
usad8 r4, r7, lr ; calculate sum of positive differences
usad8 r5, r6, lr ; calculate sum of negative differences
orr r6, r6, r7 ; differences of all 4 pixels
; calculate total sum
add r8, r8, r4 ; add positive differences to sum
sub r8, r8, r5 ; subtract negative differences from sum
; calculate sse
uxtb16 r5, r6 ; byte (two pixels) to halfwords
uxtb16 r10, r6, ror #8 ; another two pixels to halfwords
smlad r11, r5, r5, r11 ; dual signed multiply, add and accumulate (1)
smlad r11, r10, r10, r11 ; dual signed multiply, add and accumulate (2)
subs r12, r12, #1
bne loop
; return stuff
ldr r6, [sp, #40] ; get address of sse
mul r0, r8, r8 ; sum * sum
str r11, [r6] ; store sse
sub r0, r11, r0, lsr #8 ; return (sse - ((sum * sum) >> 8))
ldmfd sp!, {r4-r12, pc}
ENDP
END

101
media/libvpx/vp8/common/arm/armv6/vp8_variance8x8_armv6.asm
Просмотреть файл

@ -1,101 +0,0 @@
;
; Copyright (c) 2011 The WebM project authors. All Rights Reserved.
;
; Use of this source code is governed by a BSD-style license
; that can be found in the LICENSE file in the root of the source
; tree. An additional intellectual property rights grant can be found
; in the file PATENTS. All contributing project authors may
; be found in the AUTHORS file in the root of the source tree.
;
EXPORT |vp8_variance8x8_armv6|
ARM
AREA ||.text||, CODE, READONLY, ALIGN=2
; r0 unsigned char *src_ptr
; r1 int source_stride
; r2 unsigned char *ref_ptr
; r3 int recon_stride
; stack unsigned int *sse
|vp8_variance8x8_armv6| PROC
push {r4-r10, lr}
pld [r0, r1, lsl #0]
pld [r2, r3, lsl #0]
mov r12, #8 ; set loop counter to 8 (=block height)
mov r4, #0 ; initialize sum = 0
mov r5, #0 ; initialize sse = 0
loop
; 1st 4 pixels
ldr r6, [r0, #0x0] ; load 4 src pixels
ldr r7, [r2, #0x0] ; load 4 ref pixels
mov lr, #0 ; constant zero
usub8 r8, r6, r7 ; calculate difference
pld [r0, r1, lsl #1]
sel r10, r8, lr ; select bytes with positive difference
usub8 r9, r7, r6 ; calculate difference with reversed operands
pld [r2, r3, lsl #1]
sel r8, r9, lr ; select bytes with negative difference
; calculate partial sums
usad8 r6, r10, lr ; calculate sum of positive differences
usad8 r7, r8, lr ; calculate sum of negative differences
orr r8, r8, r10 ; differences of all 4 pixels
; calculate total sum
add r4, r4, r6 ; add positive differences to sum
sub r4, r4, r7 ; subtract negative differences from sum
; calculate sse
uxtb16 r7, r8 ; byte (two pixels) to halfwords
uxtb16 r10, r8, ror #8 ; another two pixels to halfwords
smlad r5, r7, r7, r5 ; dual signed multiply, add and accumulate (1)
; 2nd 4 pixels
ldr r6, [r0, #0x4] ; load 4 src pixels
ldr r7, [r2, #0x4] ; load 4 ref pixels
smlad r5, r10, r10, r5 ; dual signed multiply, add and accumulate (2)
usub8 r8, r6, r7 ; calculate difference
add r0, r0, r1 ; set src_ptr to next row
sel r10, r8, lr ; select bytes with positive difference
usub8 r9, r7, r6 ; calculate difference with reversed operands
add r2, r2, r3 ; set dst_ptr to next row
sel r8, r9, lr ; select bytes with negative difference
; calculate partial sums
usad8 r6, r10, lr ; calculate sum of positive differences
usad8 r7, r8, lr ; calculate sum of negative differences
orr r8, r8, r10 ; differences of all 4 pixels
; calculate total sum
add r4, r4, r6 ; add positive differences to sum
sub r4, r4, r7 ; subtract negative differences from sum
; calculate sse
uxtb16 r7, r8 ; byte (two pixels) to halfwords
uxtb16 r10, r8, ror #8 ; another two pixels to halfwords
smlad r5, r7, r7, r5 ; dual signed multiply, add and accumulate (1)
subs r12, r12, #1 ; next row
smlad r5, r10, r10, r5 ; dual signed multiply, add and accumulate (2)
bne loop
; return stuff
ldr r8, [sp, #32] ; get address of sse
mul r1, r4, r4 ; sum * sum
str r5, [r8] ; store sse
sub r0, r5, r1, ASR #6 ; return (sse - ((sum * sum) >> 6))
pop {r4-r10, pc}
ENDP
END

6
media/libvpx/vp8/common/arm/filter_arm.c
Просмотреть файл

@ -99,7 +99,7 @@ void vp8_sixtap_predict4x4_armv6
{
const short *HFilter;
const short *VFilter;
DECLARE_ALIGNED_ARRAY(4, short, FData, 12*4); /* Temp data buffer used in filtering */
DECLARE_ALIGNED(4, short, FData[12*4]); /* Temp data buffer used in filtering */
HFilter = vp8_sub_pel_filters[xoffset]; /* 6 tap */
@ -147,7 +147,7 @@ void vp8_sixtap_predict8x8_armv6
{
const short *HFilter;
const short *VFilter;
DECLARE_ALIGNED_ARRAY(4, short, FData, 16*8); /* Temp data buffer used in filtering */
DECLARE_ALIGNED(4, short, FData[16*8]); /* Temp data buffer used in filtering */
HFilter = vp8_sub_pel_filters[xoffset]; /* 6 tap */
VFilter = vp8_sub_pel_filters[yoffset]; /* 6 tap */
@ -189,7 +189,7 @@ void vp8_sixtap_predict16x16_armv6
{
const short *HFilter;
const short *VFilter;
DECLARE_ALIGNED_ARRAY(4, short, FData, 24*16); /* Temp data buffer used in filtering */
DECLARE_ALIGNED(4, short, FData[24*16]); /* Temp data buffer used in filtering */
HFilter = vp8_sub_pel_filters[xoffset]; /* 6 tap */
VFilter = vp8_sub_pel_filters[yoffset]; /* 6 tap */

184
media/libvpx/vp8/common/arm/neon/sad_neon.c
Просмотреть файл

@ -1,184 +0,0 @@
/*
* Copyright (c) 2014 The WebM project authors. All Rights Reserved.
*
* Use of this source code is governed by a BSD-style license
* that can be found in the LICENSE file in the root of the source
* tree. An additional intellectual property rights grant can be found
* in the file PATENTS. All contributing project authors may
* be found in the AUTHORS file in the root of the source tree.
*/
#include <arm_neon.h>
unsigned int vp8_sad8x8_neon(
unsigned char *src_ptr,
int src_stride,
unsigned char *ref_ptr,
int ref_stride) {
uint8x8_t d0, d8;
uint16x8_t q12;
uint32x4_t q1;
uint64x2_t q3;
uint32x2_t d5;
int i;
d0 = vld1_u8(src_ptr);
src_ptr += src_stride;
d8 = vld1_u8(ref_ptr);
ref_ptr += ref_stride;
q12 = vabdl_u8(d0, d8);
for (i = 0; i < 7; i++) {
d0 = vld1_u8(src_ptr);
src_ptr += src_stride;
d8 = vld1_u8(ref_ptr);
ref_ptr += ref_stride;
q12 = vabal_u8(q12, d0, d8);
}
q1 = vpaddlq_u16(q12);
q3 = vpaddlq_u32(q1);
d5 = vadd_u32(vreinterpret_u32_u64(vget_low_u64(q3)),
vreinterpret_u32_u64(vget_high_u64(q3)));
return vget_lane_u32(d5, 0);
}
unsigned int vp8_sad8x16_neon(
unsigned char *src_ptr,
int src_stride,
unsigned char *ref_ptr,
int ref_stride) {
uint8x8_t d0, d8;
uint16x8_t q12;
uint32x4_t q1;
uint64x2_t q3;
uint32x2_t d5;
int i;
d0 = vld1_u8(src_ptr);
src_ptr += src_stride;
d8 = vld1_u8(ref_ptr);
ref_ptr += ref_stride;
q12 = vabdl_u8(d0, d8);
for (i = 0; i < 15; i++) {
d0 = vld1_u8(src_ptr);
src_ptr += src_stride;
d8 = vld1_u8(ref_ptr);
ref_ptr += ref_stride;
q12 = vabal_u8(q12, d0, d8);
}
q1 = vpaddlq_u16(q12);
q3 = vpaddlq_u32(q1);
d5 = vadd_u32(vreinterpret_u32_u64(vget_low_u64(q3)),
vreinterpret_u32_u64(vget_high_u64(q3)));
return vget_lane_u32(d5, 0);
}
unsigned int vp8_sad4x4_neon(
unsigned char *src_ptr,
int src_stride,
unsigned char *ref_ptr,
int ref_stride) {
uint8x8_t d0, d8;
uint16x8_t q12;
uint32x2_t d1;
uint64x1_t d3;
int i;
d0 = vld1_u8(src_ptr);
src_ptr += src_stride;
d8 = vld1_u8(ref_ptr);
ref_ptr += ref_stride;
q12 = vabdl_u8(d0, d8);
for (i = 0; i < 3; i++) {
d0 = vld1_u8(src_ptr);
src_ptr += src_stride;
d8 = vld1_u8(ref_ptr);
ref_ptr += ref_stride;
q12 = vabal_u8(q12, d0, d8);
}
d1 = vpaddl_u16(vget_low_u16(q12));
d3 = vpaddl_u32(d1);
return vget_lane_u32(vreinterpret_u32_u64(d3), 0);
}
unsigned int vp8_sad16x16_neon(
unsigned char *src_ptr,
int src_stride,
unsigned char *ref_ptr,
int ref_stride) {
uint8x16_t q0, q4;
uint16x8_t q12, q13;
uint32x4_t q1;
uint64x2_t q3;
uint32x2_t d5;
int i;
q0 = vld1q_u8(src_ptr);
src_ptr += src_stride;
q4 = vld1q_u8(ref_ptr);
ref_ptr += ref_stride;
q12 = vabdl_u8(vget_low_u8(q0), vget_low_u8(q4));
q13 = vabdl_u8(vget_high_u8(q0), vget_high_u8(q4));
for (i = 0; i < 15; i++) {
q0 = vld1q_u8(src_ptr);
src_ptr += src_stride;
q4 = vld1q_u8(ref_ptr);
ref_ptr += ref_stride;
q12 = vabal_u8(q12, vget_low_u8(q0), vget_low_u8(q4));
q13 = vabal_u8(q13, vget_high_u8(q0), vget_high_u8(q4));
}
q12 = vaddq_u16(q12, q13);
q1 = vpaddlq_u16(q12);
q3 = vpaddlq_u32(q1);
d5 = vadd_u32(vreinterpret_u32_u64(vget_low_u64(q3)),
vreinterpret_u32_u64(vget_high_u64(q3)));
return vget_lane_u32(d5, 0);
}
unsigned int vp8_sad16x8_neon(
unsigned char *src_ptr,
int src_stride,
unsigned char *ref_ptr,
int ref_stride) {
uint8x16_t q0, q4;
uint16x8_t q12, q13;
uint32x4_t q1;
uint64x2_t q3;
uint32x2_t d5;
int i;
q0 = vld1q_u8(src_ptr);
src_ptr += src_stride;
q4 = vld1q_u8(ref_ptr);
ref_ptr += ref_stride;
q12 = vabdl_u8(vget_low_u8(q0), vget_low_u8(q4));
q13 = vabdl_u8(vget_high_u8(q0), vget_high_u8(q4));
for (i = 0; i < 7; i++) {
q0 = vld1q_u8(src_ptr);
src_ptr += src_stride;
q4 = vld1q_u8(ref_ptr);
ref_ptr += ref_stride;
q12 = vabal_u8(q12, vget_low_u8(q0), vget_low_u8(q4));
q13 = vabal_u8(q13, vget_high_u8(q0), vget_high_u8(q4));
}
q12 = vaddq_u16(q12, q13);
q1 = vpaddlq_u16(q12);
q3 = vpaddlq_u32(q1);
d5 = vadd_u32(vreinterpret_u32_u64(vget_low_u64(q3)),
vreinterpret_u32_u64(vget_high_u64(q3)));
return vget_lane_u32(d5, 0);
}

320
media/libvpx/vp8/common/arm/neon/variance_neon.c
Просмотреть файл

@ -1,320 +0,0 @@
/*
* Copyright (c) 2014 The WebM project authors. All Rights Reserved.
*
* Use of this source code is governed by a BSD-style license
* that can be found in the LICENSE file in the root of the source
* tree. An additional intellectual property rights grant can be found
* in the file PATENTS. All contributing project authors may
* be found in the AUTHORS file in the root of the source tree.
*/
#include <arm_neon.h>
#include "vpx_ports/mem.h"
unsigned int vp8_variance16x16_neon(
const unsigned char *src_ptr,
int source_stride,
const unsigned char *ref_ptr,
int recon_stride,
unsigned int *sse) {
int i;
int16x4_t d22s16, d23s16, d24s16, d25s16, d26s16, d27s16, d28s16, d29s16;
uint32x2_t d0u32, d10u32;
int64x1_t d0s64, d1s64;
uint8x16_t q0u8, q1u8, q2u8, q3u8;
uint16x8_t q11u16, q12u16, q13u16, q14u16;
int32x4_t q8s32, q9s32, q10s32;
int64x2_t q0s64, q1s64, q5s64;
q8s32 = vdupq_n_s32(0);
q9s32 = vdupq_n_s32(0);
q10s32 = vdupq_n_s32(0);
for (i = 0; i < 8; i++) {
q0u8 = vld1q_u8(src_ptr);
src_ptr += source_stride;
q1u8 = vld1q_u8(src_ptr);
src_ptr += source_stride;
__builtin_prefetch(src_ptr);
q2u8 = vld1q_u8(ref_ptr);
ref_ptr += recon_stride;
q3u8 = vld1q_u8(ref_ptr);
ref_ptr += recon_stride;
__builtin_prefetch(ref_ptr);
q11u16 = vsubl_u8(vget_low_u8(q0u8), vget_low_u8(q2u8));
q12u16 = vsubl_u8(vget_high_u8(q0u8), vget_high_u8(q2u8));
q13u16 = vsubl_u8(vget_low_u8(q1u8), vget_low_u8(q3u8));
q14u16 = vsubl_u8(vget_high_u8(q1u8), vget_high_u8(q3u8));
d22s16 = vreinterpret_s16_u16(vget_low_u16(q11u16));
d23s16 = vreinterpret_s16_u16(vget_high_u16(q11u16));
q8s32 = vpadalq_s16(q8s32, vreinterpretq_s16_u16(q11u16));
q9s32 = vmlal_s16(q9s32, d22s16, d22s16);
q10s32 = vmlal_s16(q10s32, d23s16, d23s16);
d24s16 = vreinterpret_s16_u16(vget_low_u16(q12u16));
d25s16 = vreinterpret_s16_u16(vget_high_u16(q12u16));
q8s32 = vpadalq_s16(q8s32, vreinterpretq_s16_u16(q12u16));
q9s32 = vmlal_s16(q9s32, d24s16, d24s16);
q10s32 = vmlal_s16(q10s32, d25s16, d25s16);
d26s16 = vreinterpret_s16_u16(vget_low_u16(q13u16));
d27s16 = vreinterpret_s16_u16(vget_high_u16(q13u16));
q8s32 = vpadalq_s16(q8s32, vreinterpretq_s16_u16(q13u16));
q9s32 = vmlal_s16(q9s32, d26s16, d26s16);
q10s32 = vmlal_s16(q10s32, d27s16, d27s16);
d28s16 = vreinterpret_s16_u16(vget_low_u16(q14u16));
d29s16 = vreinterpret_s16_u16(vget_high_u16(q14u16));
q8s32 = vpadalq_s16(q8s32, vreinterpretq_s16_u16(q14u16));
q9s32 = vmlal_s16(q9s32, d28s16, d28s16);
q10s32 = vmlal_s16(q10s32, d29s16, d29s16);
}
q10s32 = vaddq_s32(q10s32, q9s32);
q0s64 = vpaddlq_s32(q8s32);
q1s64 = vpaddlq_s32(q10s32);
d0s64 = vadd_s64(vget_low_s64(q0s64), vget_high_s64(q0s64));
d1s64 = vadd_s64(vget_low_s64(q1s64), vget_high_s64(q1s64));
q5s64 = vmull_s32(vreinterpret_s32_s64(d0s64),
vreinterpret_s32_s64(d0s64));
vst1_lane_u32((uint32_t *)sse, vreinterpret_u32_s64(d1s64), 0);
d10u32 = vshr_n_u32(vreinterpret_u32_s64(vget_low_s64(q5s64)), 8);
d0u32 = vsub_u32(vreinterpret_u32_s64(d1s64), d10u32);
return vget_lane_u32(d0u32, 0);
}
unsigned int vp8_variance16x8_neon(
const unsigned char *src_ptr,
int source_stride,
const unsigned char *ref_ptr,
int recon_stride,
unsigned int *sse) {
int i;
int16x4_t d22s16, d23s16, d24s16, d25s16, d26s16, d27s16, d28s16, d29s16;
uint32x2_t d0u32, d10u32;
int64x1_t d0s64, d1s64;
uint8x16_t q0u8, q1u8, q2u8, q3u8;
uint16x8_t q11u16, q12u16, q13u16, q14u16;
int32x4_t q8s32, q9s32, q10s32;
int64x2_t q0s64, q1s64, q5s64;
q8s32 = vdupq_n_s32(0);
q9s32 = vdupq_n_s32(0);
q10s32 = vdupq_n_s32(0);
for (i = 0; i < 4; i++) { // variance16x8_neon_loop
q0u8 = vld1q_u8(src_ptr);
src_ptr += source_stride;
q1u8 = vld1q_u8(src_ptr);
src_ptr += source_stride;
__builtin_prefetch(src_ptr);
q2u8 = vld1q_u8(ref_ptr);
ref_ptr += recon_stride;
q3u8 = vld1q_u8(ref_ptr);
ref_ptr += recon_stride;
__builtin_prefetch(ref_ptr);
q11u16 = vsubl_u8(vget_low_u8(q0u8), vget_low_u8(q2u8));
q12u16 = vsubl_u8(vget_high_u8(q0u8), vget_high_u8(q2u8));
q13u16 = vsubl_u8(vget_low_u8(q1u8), vget_low_u8(q3u8));
q14u16 = vsubl_u8(vget_high_u8(q1u8), vget_high_u8(q3u8));
d22s16 = vreinterpret_s16_u16(vget_low_u16(q11u16));
d23s16 = vreinterpret_s16_u16(vget_high_u16(q11u16));
q8s32 = vpadalq_s16(q8s32, vreinterpretq_s16_u16(q11u16));
q9s32 = vmlal_s16(q9s32, d22s16, d22s16);
q10s32 = vmlal_s16(q10s32, d23s16, d23s16);
d24s16 = vreinterpret_s16_u16(vget_low_u16(q12u16));
d25s16 = vreinterpret_s16_u16(vget_high_u16(q12u16));
q8s32 = vpadalq_s16(q8s32, vreinterpretq_s16_u16(q12u16));
q9s32 = vmlal_s16(q9s32, d24s16, d24s16);
q10s32 = vmlal_s16(q10s32, d25s16, d25s16);
d26s16 = vreinterpret_s16_u16(vget_low_u16(q13u16));
d27s16 = vreinterpret_s16_u16(vget_high_u16(q13u16));
q8s32 = vpadalq_s16(q8s32, vreinterpretq_s16_u16(q13u16));
q9s32 = vmlal_s16(q9s32, d26s16, d26s16);
q10s32 = vmlal_s16(q10s32, d27s16, d27s16);
d28s16 = vreinterpret_s16_u16(vget_low_u16(q14u16));
d29s16 = vreinterpret_s16_u16(vget_high_u16(q14u16));
q8s32 = vpadalq_s16(q8s32, vreinterpretq_s16_u16(q14u16));
q9s32 = vmlal_s16(q9s32, d28s16, d28s16);
q10s32 = vmlal_s16(q10s32, d29s16, d29s16);
}
q10s32 = vaddq_s32(q10s32, q9s32);
q0s64 = vpaddlq_s32(q8s32);
q1s64 = vpaddlq_s32(q10s32);
d0s64 = vadd_s64(vget_low_s64(q0s64), vget_high_s64(q0s64));
d1s64 = vadd_s64(vget_low_s64(q1s64), vget_high_s64(q1s64));
q5s64 = vmull_s32(vreinterpret_s32_s64(d0s64),
vreinterpret_s32_s64(d0s64));
vst1_lane_u32((uint32_t *)sse, vreinterpret_u32_s64(d1s64), 0);
d10u32 = vshr_n_u32(vreinterpret_u32_s64(vget_low_s64(q5s64)), 7);
d0u32 = vsub_u32(vreinterpret_u32_s64(d1s64), d10u32);
return vget_lane_u32(d0u32, 0);
}
unsigned int vp8_variance8x16_neon(
const unsigned char *src_ptr,
int source_stride,
const unsigned char *ref_ptr,
int recon_stride,
unsigned int *sse) {
int i;
uint8x8_t d0u8, d2u8, d4u8, d6u8;
int16x4_t d22s16, d23s16, d24s16, d25s16;
uint32x2_t d0u32, d10u32;
int64x1_t d0s64, d1s64;
uint16x8_t q11u16, q12u16;
int32x4_t q8s32, q9s32, q10s32;
int64x2_t q0s64, q1s64, q5s64;
q8s32 = vdupq_n_s32(0);
q9s32 = vdupq_n_s32(0);
q10s32 = vdupq_n_s32(0);
for (i = 0; i < 8; i++) { // variance8x16_neon_loop
d0u8 = vld1_u8(src_ptr);
src_ptr += source_stride;
d2u8 = vld1_u8(src_ptr);
src_ptr += source_stride;
__builtin_prefetch(src_ptr);
d4u8 = vld1_u8(ref_ptr);
ref_ptr += recon_stride;
d6u8 = vld1_u8(ref_ptr);
ref_ptr += recon_stride;
__builtin_prefetch(ref_ptr);
q11u16 = vsubl_u8(d0u8, d4u8);
q12u16 = vsubl_u8(d2u8, d6u8);
d22s16 = vreinterpret_s16_u16(vget_low_u16(q11u16));
d23s16 = vreinterpret_s16_u16(vget_high_u16(q11u16));
q8s32 = vpadalq_s16(q8s32, vreinterpretq_s16_u16(q11u16));
q9s32 = vmlal_s16(q9s32, d22s16, d22s16);
q10s32 = vmlal_s16(q10s32, d23s16, d23s16);
d24s16 = vreinterpret_s16_u16(vget_low_u16(q12u16));
d25s16 = vreinterpret_s16_u16(vget_high_u16(q12u16));
q8s32 = vpadalq_s16(q8s32, vreinterpretq_s16_u16(q12u16));
q9s32 = vmlal_s16(q9s32, d24s16, d24s16);
q10s32 = vmlal_s16(q10s32, d25s16, d25s16);
}
q10s32 = vaddq_s32(q10s32, q9s32);
q0s64 = vpaddlq_s32(q8s32);
q1s64 = vpaddlq_s32(q10s32);
d0s64 = vadd_s64(vget_low_s64(q0s64), vget_high_s64(q0s64));
d1s64 = vadd_s64(vget_low_s64(q1s64), vget_high_s64(q1s64));
q5s64 = vmull_s32(vreinterpret_s32_s64(d0s64),
vreinterpret_s32_s64(d0s64));
vst1_lane_u32((uint32_t *)sse, vreinterpret_u32_s64(d1s64), 0);
d10u32 = vshr_n_u32(vreinterpret_u32_s64(vget_low_s64(q5s64)), 7);
d0u32 = vsub_u32(vreinterpret_u32_s64(d1s64), d10u32);
return vget_lane_u32(d0u32, 0);
}
unsigned int vp8_variance8x8_neon(
const unsigned char *src_ptr,
int source_stride,
const unsigned char *ref_ptr,
int recon_stride,
unsigned int *sse) {
int i;
uint8x8_t d0u8, d1u8, d2u8, d3u8, d4u8, d5u8, d6u8, d7u8;
int16x4_t d22s16, d23s16, d24s16, d25s16, d26s16, d27s16, d28s16, d29s16;
uint32x2_t d0u32, d10u32;
int64x1_t d0s64, d1s64;
uint16x8_t q11u16, q12u16, q13u16, q14u16;
int32x4_t q8s32, q9s32, q10s32;
int64x2_t q0s64, q1s64, q5s64;
q8s32 = vdupq_n_s32(0);
q9s32 = vdupq_n_s32(0);
q10s32 = vdupq_n_s32(0);
for (i = 0; i < 2; i++) { // variance8x8_neon_loop
d0u8 = vld1_u8(src_ptr);
src_ptr += source_stride;
d1u8 = vld1_u8(src_ptr);
src_ptr += source_stride;
d2u8 = vld1_u8(src_ptr);
src_ptr += source_stride;
d3u8 = vld1_u8(src_ptr);
src_ptr += source_stride;
d4u8 = vld1_u8(ref_ptr);
ref_ptr += recon_stride;
d5u8 = vld1_u8(ref_ptr);
ref_ptr += recon_stride;
d6u8 = vld1_u8(ref_ptr);
ref_ptr += recon_stride;
d7u8 = vld1_u8(ref_ptr);
ref_ptr += recon_stride;
q11u16 = vsubl_u8(d0u8, d4u8);
q12u16 = vsubl_u8(d1u8, d5u8);
q13u16 = vsubl_u8(d2u8, d6u8);
q14u16 = vsubl_u8(d3u8, d7u8);
d22s16 = vreinterpret_s16_u16(vget_low_u16(q11u16));
d23s16 = vreinterpret_s16_u16(vget_high_u16(q11u16));
q8s32 = vpadalq_s16(q8s32, vreinterpretq_s16_u16(q11u16));
q9s32 = vmlal_s16(q9s32, d22s16, d22s16);
q10s32 = vmlal_s16(q10s32, d23s16, d23s16);
d24s16 = vreinterpret_s16_u16(vget_low_u16(q12u16));
d25s16 = vreinterpret_s16_u16(vget_high_u16(q12u16));
q8s32 = vpadalq_s16(q8s32, vreinterpretq_s16_u16(q12u16));
q9s32 = vmlal_s16(q9s32, d24s16, d24s16);
q10s32 = vmlal_s16(q10s32, d25s16, d25s16);
d26s16 = vreinterpret_s16_u16(vget_low_u16(q13u16));
d27s16 = vreinterpret_s16_u16(vget_high_u16(q13u16));
q8s32 = vpadalq_s16(q8s32, vreinterpretq_s16_u16(q13u16));
q9s32 = vmlal_s16(q9s32, d26s16, d26s16);
q10s32 = vmlal_s16(q10s32, d27s16, d27s16);
d28s16 = vreinterpret_s16_u16(vget_low_u16(q14u16));
d29s16 = vreinterpret_s16_u16(vget_high_u16(q14u16));
q8s32 = vpadalq_s16(q8s32, vreinterpretq_s16_u16(q14u16));
q9s32 = vmlal_s16(q9s32, d28s16, d28s16);
q10s32 = vmlal_s16(q10s32, d29s16, d29s16);
}
q10s32 = vaddq_s32(q10s32, q9s32);
q0s64 = vpaddlq_s32(q8s32);
q1s64 = vpaddlq_s32(q10s32);
d0s64 = vadd_s64(vget_low_s64(q0s64), vget_high_s64(q0s64));
d1s64 = vadd_s64(vget_low_s64(q1s64), vget_high_s64(q1s64));
q5s64 = vmull_s32(vreinterpret_s32_s64(d0s64),
vreinterpret_s32_s64(d0s64));
vst1_lane_u32((uint32_t *)sse, vreinterpret_u32_s64(d1s64), 0);
d10u32 = vshr_n_u32(vreinterpret_u32_s64(vget_low_s64(q5s64)), 6);
d0u32 = vsub_u32(vreinterpret_u32_s64(d1s64), d10u32);
return vget_lane_u32(d0u32, 0);
}

34
media/libvpx/vp8/common/arm/neon/vp8_subpixelvariance_neon.c
Просмотреть файл

@ -32,7 +32,7 @@ unsigned int vp8_sub_pixel_variance16x16_neon_func(
int dst_pixels_per_line,
unsigned int *sse) {
int i;
DECLARE_ALIGNED_ARRAY(16, unsigned char, tmp, 528);
DECLARE_ALIGNED(16, unsigned char, tmp[528]);
unsigned char *tmpp;
unsigned char *tmpp2;
uint8x8_t d0u8, d1u8, d2u8, d3u8, d4u8, d5u8, d6u8, d7u8, d8u8, d9u8;
@ -911,12 +911,6 @@ unsigned int vp8_variance_halfpixvar16x16_hv_neon(
return vget_lane_u32(d0u32, 0);
}
enum { kWidth8 = 8 };
enum { kHeight8 = 8 };
enum { kHeight8PlusOne = 9 };
enum { kPixelStepOne = 1 };
enum { kAlign16 = 16 };
#define FILTER_BITS 7
static INLINE int horizontal_add_s16x8(const int16x8_t v_16x8) {
@ -968,8 +962,8 @@ static unsigned int variance8x8_neon(const uint8_t *a, int a_stride,
const uint8_t *b, int b_stride,
unsigned int *sse) {
int sum;
variance_neon_w8(a, a_stride, b, b_stride, kWidth8, kHeight8, sse, &sum);
return *sse - (((int64_t)sum * sum) / (kWidth8 * kHeight8));
variance_neon_w8(a, a_stride, b, b_stride, 8, 8, sse, &sum);
return *sse - (((int64_t)sum * sum) / (8 * 8));
}
static void var_filter_block2d_bil_w8(const uint8_t *src_ptr,
@ -1003,21 +997,21 @@ unsigned int vp8_sub_pixel_variance8x8_neon(
const unsigned char *dst,
int dst_stride,
unsigned int *sse) {
DECLARE_ALIGNED_ARRAY(kAlign16, uint8_t, temp2, kHeight8PlusOne * kWidth8);
DECLARE_ALIGNED_ARRAY(kAlign16, uint8_t, fdata3, kHeight8PlusOne * kWidth8);
DECLARE_ALIGNED(16, uint8_t, temp2[9 * 8]);
DECLARE_ALIGNED(16, uint8_t, fdata3[9 * 8]);
if (xoffset == 0) {
var_filter_block2d_bil_w8(src, temp2, src_stride, kWidth8, kHeight8,
kWidth8, bilinear_taps_coeff[yoffset]);
var_filter_block2d_bil_w8(src, temp2, src_stride, 8, 8,
8, bilinear_taps_coeff[yoffset]);
} else if (yoffset == 0) {
var_filter_block2d_bil_w8(src, temp2, src_stride, kPixelStepOne,
kHeight8PlusOne, kWidth8,
var_filter_block2d_bil_w8(src, temp2, src_stride, 1,
9, 8,
bilinear_taps_coeff[xoffset]);
} else {
var_filter_block2d_bil_w8(src, fdata3, src_stride, kPixelStepOne,
kHeight8PlusOne, kWidth8,
var_filter_block2d_bil_w8(src, fdata3, src_stride, 1,
9, 8,
bilinear_taps_coeff[xoffset]);
var_filter_block2d_bil_w8(fdata3, temp2, kWidth8, kWidth8, kHeight8,
kWidth8, bilinear_taps_coeff[yoffset]);
var_filter_block2d_bil_w8(fdata3, temp2, 8, 8, 8,
8, bilinear_taps_coeff[yoffset]);
}
return variance8x8_neon(temp2, kWidth8, dst, dst_stride, sse);
return variance8x8_neon(temp2, 8, dst, dst_stride, sse);
}

19
media/libvpx/vp8/common/arm/variance_arm.c
Просмотреть файл

@ -9,10 +9,14 @@
*/
#include "vpx_config.h"
#include "vp8_rtcd.h"
#include "./vp8_rtcd.h"
#include "./vpx_dsp_rtcd.h"
#include "vp8/common/variance.h"
#include "vp8/common/filter.h"
// TODO(johannkoenig): Move this to vpx_dsp or vp8/encoder
#if CONFIG_VP8_ENCODER
#if HAVE_MEDIA
#include "vp8/common/arm/bilinearfilter_arm.h"
@ -40,8 +44,8 @@ unsigned int vp8_sub_pixel_variance8x8_armv6
vp8_filter_block2d_bil_second_pass_armv6(first_pass, second_pass,
8, 8, 8, VFilter);
return vp8_variance8x8_armv6(second_pass, 8, dst_ptr,
dst_pixels_per_line, sse);
return vpx_variance8x8_media(second_pass, 8, dst_ptr,
dst_pixels_per_line, sse);
}
unsigned int vp8_sub_pixel_variance16x16_armv6
@ -86,13 +90,13 @@ unsigned int vp8_sub_pixel_variance16x16_armv6
vp8_filter_block2d_bil_second_pass_armv6(first_pass, second_pass,
16, 16, 16, VFilter);
var = vp8_variance16x16_armv6(second_pass, 16, dst_ptr,
dst_pixels_per_line, sse);
var = vpx_variance16x16_media(second_pass, 16, dst_ptr,
dst_pixels_per_line, sse);
}
return var;
}
#endif /* HAVE_MEDIA */
#endif // HAVE_MEDIA
#if HAVE_NEON
@ -129,4 +133,5 @@ unsigned int vp8_sub_pixel_variance16x16_neon
return vp8_sub_pixel_variance16x16_neon_func(src_ptr, src_pixels_per_line, xoffset, yoffset, dst_ptr, dst_pixels_per_line, sse);
}
#endif
#endif // HAVE_NEON
#endif // CONFIG_VP8_ENCODER

8
media/libvpx/vp8/common/common.h
Просмотреть файл

@ -29,19 +29,19 @@ extern "C" {
#define vp8_copy( Dest, Src) { \
assert( sizeof( Dest) == sizeof( Src)); \
vpx_memcpy( Dest, Src, sizeof( Src)); \
memcpy( Dest, Src, sizeof( Src)); \
}
/* Use this for variably-sized arrays. */
#define vp8_copy_array( Dest, Src, N) { \
assert( sizeof( *Dest) == sizeof( *Src)); \
vpx_memcpy( Dest, Src, N * sizeof( *Src)); \
memcpy( Dest, Src, N * sizeof( *Src)); \
}
#define vp8_zero( Dest) vpx_memset( &Dest, 0, sizeof( Dest));
#define vp8_zero( Dest) memset( &Dest, 0, sizeof( Dest));
#define vp8_zero_array( Dest, N) vpx_memset( Dest, 0, N * sizeof( *Dest));
#define vp8_zero_array( Dest, N) memset( Dest, 0, N * sizeof( *Dest));
#ifdef __cplusplus

32
media/libvpx/vp8/common/copy_c.c Normal file
Просмотреть файл

@ -0,0 +1,32 @@
/*
* Copyright (c) 2010 The WebM project authors. All Rights Reserved.
*
* Use of this source code is governed by a BSD-style license
* that can be found in the LICENSE file in the root of the source
* tree. An additional intellectual property rights grant can be found
* in the file PATENTS. All contributing project authors may
* be found in the AUTHORS file in the root of the source tree.
*/
#include <string.h>
#include "./vp8_rtcd.h"
#include "vpx/vpx_integer.h"
/* Copy 2 macroblocks to a buffer */
void vp8_copy32xn_c(const unsigned char *src_ptr, int src_stride,
unsigned char *dst_ptr, int dst_stride,
int height)
{
int r;
for (r = 0; r < height; r++)
{
memcpy(dst_ptr, src_ptr, 32);
src_ptr += src_stride;
dst_ptr += dst_stride;
}
}

2
media/libvpx/vp8/common/debugmodes.c
Просмотреть файл

@ -81,7 +81,6 @@ void vp8_print_modes_and_motion_vectors(MODE_INFO *mi, int rows, int cols, int f
fprintf(mvs, "\n");
/* print out the block modes */
mb_index = 0;
fprintf(mvs, "Mbs for Frame %d\n", frame);
{
int b_row;
@ -129,7 +128,6 @@ void vp8_print_modes_and_motion_vectors(MODE_INFO *mi, int rows, int cols, int f
/* print out the block modes */
mb_index = 0;
fprintf(mvs, "MVs for Frame %d\n", frame);
{
int b_row;

2
media/libvpx/vp8/common/dequantize.c
Просмотреть файл

@ -38,6 +38,6 @@ void vp8_dequant_idct_add_c(short *input, short *dq,
vp8_short_idct4x4llm_c(input, dest, stride, dest, stride);
vpx_memset(input, 0, 32);
memset(input, 0, 32);
}

3
media/libvpx/vp8/common/entropy.c
Просмотреть файл

@ -183,7 +183,6 @@ const vp8_extra_bit_struct vp8_extra_bits[12] =
void vp8_default_coef_probs(VP8_COMMON *pc)
{
vpx_memcpy(pc->fc.coef_probs, default_coef_probs,
sizeof(default_coef_probs));
memcpy(pc->fc.coef_probs, default_coef_probs, sizeof(default_coef_probs));
}

8
media/libvpx/vp8/common/entropymode.c
Просмотреть файл

@ -159,13 +159,13 @@ const vp8_tree_index vp8_small_mvtree [14] =
void vp8_init_mbmode_probs(VP8_COMMON *x)
{
vpx_memcpy(x->fc.ymode_prob, vp8_ymode_prob, sizeof(vp8_ymode_prob));
vpx_memcpy(x->fc.uv_mode_prob, vp8_uv_mode_prob, sizeof(vp8_uv_mode_prob));
vpx_memcpy(x->fc.sub_mv_ref_prob, sub_mv_ref_prob, sizeof(sub_mv_ref_prob));
memcpy(x->fc.ymode_prob, vp8_ymode_prob, sizeof(vp8_ymode_prob));
memcpy(x->fc.uv_mode_prob, vp8_uv_mode_prob, sizeof(vp8_uv_mode_prob));
memcpy(x->fc.sub_mv_ref_prob, sub_mv_ref_prob, sizeof(sub_mv_ref_prob));
}
void vp8_default_bmode_probs(vp8_prob p [VP8_BINTRAMODES-1])
{
vpx_memcpy(p, vp8_bmode_prob, sizeof(vp8_bmode_prob));
memcpy(p, vp8_bmode_prob, sizeof(vp8_bmode_prob));
}

10
media/libvpx/vp8/common/extend.c
Просмотреть файл

@ -40,9 +40,9 @@ static void copy_and_extend_plane
for (i = 0; i < h; i++)
{
vpx_memset(dest_ptr1, src_ptr1[0], el);
vpx_memcpy(dest_ptr1 + el, src_ptr1, w);
vpx_memset(dest_ptr2, src_ptr2[0], er);
memset(dest_ptr1, src_ptr1[0], el);
memcpy(dest_ptr1 + el, src_ptr1, w);
memset(dest_ptr2, src_ptr2[0], er);
src_ptr1 += sp;
src_ptr2 += sp;
dest_ptr1 += dp;
@ -60,13 +60,13 @@ static void copy_and_extend_plane
for (i = 0; i < et; i++)
{
vpx_memcpy(dest_ptr1, src_ptr1, linesize);
memcpy(dest_ptr1, src_ptr1, linesize);
dest_ptr1 += dp;
}
for (i = 0; i < eb; i++)
{
vpx_memcpy(dest_ptr2, src_ptr2, linesize);
memcpy(dest_ptr2, src_ptr2, linesize);
dest_ptr2 += dp;
}
}

1
media/libvpx/vp8/common/filter.c
Просмотреть файл

@ -10,6 +10,7 @@
#include "filter.h"
#include "./vp8_rtcd.h"
DECLARE_ALIGNED(16, const short, vp8_bilinear_filters[8][2]) =
{

1
media/libvpx/vp8/common/generic/systemdependent.c
Просмотреть файл

@ -17,6 +17,7 @@
#include "vpx_ports/x86.h"
#endif
#include "vp8/common/onyxc_int.h"
#include "vp8/common/systemdependent.h"
#if CONFIG_MULTITHREAD
#if HAVE_UNISTD_H && !defined(__OS2__)

6
media/libvpx/vp8/common/idct_blk.c
Просмотреть файл

@ -33,7 +33,7 @@ void vp8_dequant_idct_add_y_block_c
else
{
vp8_dc_only_idct_add_c (q[0]*dq[0], dst, stride, dst, stride);
vpx_memset(q, 0, 2 * sizeof(q[0]));
memset(q, 0, 2 * sizeof(q[0]));
}
q += 16;
@ -59,7 +59,7 @@ void vp8_dequant_idct_add_uv_block_c
else
{
vp8_dc_only_idct_add_c (q[0]*dq[0], dstu, stride, dstu, stride);
vpx_memset(q, 0, 2 * sizeof(q[0]));
memset(q, 0, 2 * sizeof(q[0]));
}
q += 16;
@ -78,7 +78,7 @@ void vp8_dequant_idct_add_uv_block_c
else
{
vp8_dc_only_idct_add_c (q[0]*dq[0], dstv, stride, dstv, stride);
vpx_memset(q, 0, 2 * sizeof(q[0]));
memset(q, 0, 2 * sizeof(q[0]));
}
q += 16;

1
media/libvpx/vp8/common/idctllm.c
Просмотреть файл

@ -8,6 +8,7 @@
* be found in the AUTHORS file in the root of the source tree.
*/
#include "./vp8_rtcd.h"
/****************************************************************************
* Notes:

13
media/libvpx/vp8/common/loopfilter.c
Просмотреть файл

@ -82,11 +82,10 @@ void vp8_loop_filter_update_sharpness(loop_filter_info_n *lfi,
if (block_inside_limit < 1)
block_inside_limit = 1;
vpx_memset(lfi->lim[i], block_inside_limit, SIMD_WIDTH);
vpx_memset(lfi->blim[i], (2 * filt_lvl + block_inside_limit),
SIMD_WIDTH);
vpx_memset(lfi->mblim[i], (2 * (filt_lvl + 2) + block_inside_limit),
SIMD_WIDTH);
memset(lfi->lim[i], block_inside_limit, SIMD_WIDTH);
memset(lfi->blim[i], (2 * filt_lvl + block_inside_limit), SIMD_WIDTH);
memset(lfi->mblim[i], (2 * (filt_lvl + 2) + block_inside_limit),
SIMD_WIDTH);
}
}
@ -105,7 +104,7 @@ void vp8_loop_filter_init(VP8_COMMON *cm)
/* init hev threshold const vectors */
for(i = 0; i < 4 ; i++)
{
vpx_memset(lfi->hev_thr[i], i, SIMD_WIDTH);
memset(lfi->hev_thr[i], i, SIMD_WIDTH);
}
}
@ -151,7 +150,7 @@ void vp8_loop_filter_frame_init(VP8_COMMON *cm,
/* we could get rid of this if we assume that deltas are set to
* zero when not in use; encoder always uses deltas
*/
vpx_memset(lfi->lvl[seg][0], lvl_seg, 4 * 4 );
memset(lfi->lvl[seg][0], lvl_seg, 4 * 4 );
continue;
}

47
media/libvpx/vp8/common/mfqe.c
Просмотреть файл

@ -17,10 +17,11 @@
* higher quality.
*/
#include "postproc.h"
#include "variance.h"
#include "./vp8_rtcd.h"
#include "./vpx_dsp_rtcd.h"
#include "vp8/common/postproc.h"
#include "vp8/common/variance.h"
#include "vpx_mem/vpx_mem.h"
#include "vp8_rtcd.h"
#include "vpx_scale/yv12config.h"
#include <limits.h>
@ -150,36 +151,36 @@ static void multiframe_quality_enhance_block
if (blksize == 16)
{
actd = (vp8_variance16x16(yd, yd_stride, VP8_ZEROS, 0, &sse)+128)>>8;
act = (vp8_variance16x16(y, y_stride, VP8_ZEROS, 0, &sse)+128)>>8;
actd = (vpx_variance16x16(yd, yd_stride, VP8_ZEROS, 0, &sse)+128)>>8;
act = (vpx_variance16x16(y, y_stride, VP8_ZEROS, 0, &sse)+128)>>8;
#ifdef USE_SSD
sad = (vp8_variance16x16(y, y_stride, yd, yd_stride, &sse));
vpx_variance16x16(y, y_stride, yd, yd_stride, &sse);
sad = (sse + 128)>>8;
usad = (vp8_variance8x8(u, uv_stride, ud, uvd_stride, &sse));
vpx_variance8x8(u, uv_stride, ud, uvd_stride, &sse);
usad = (sse + 32)>>6;
vsad = (vp8_variance8x8(v, uv_stride, vd, uvd_stride, &sse));
vpx_variance8x8(v, uv_stride, vd, uvd_stride, &sse);
vsad = (sse + 32)>>6;
#else
sad = (vp8_sad16x16(y, y_stride, yd, yd_stride, UINT_MAX) + 128) >> 8;
usad = (vp8_sad8x8(u, uv_stride, ud, uvd_stride, UINT_MAX) + 32) >> 6;
vsad = (vp8_sad8x8(v, uv_stride, vd, uvd_stride, UINT_MAX)+ 32) >> 6;
sad = (vpx_sad16x16(y, y_stride, yd, yd_stride) + 128) >> 8;
usad = (vpx_sad8x8(u, uv_stride, ud, uvd_stride) + 32) >> 6;
vsad = (vpx_sad8x8(v, uv_stride, vd, uvd_stride)+ 32) >> 6;
#endif
}
else /* if (blksize == 8) */
{
actd = (vp8_variance8x8(yd, yd_stride, VP8_ZEROS, 0, &sse)+32)>>6;
act = (vp8_variance8x8(y, y_stride, VP8_ZEROS, 0, &sse)+32)>>6;
actd = (vpx_variance8x8(yd, yd_stride, VP8_ZEROS, 0, &sse)+32)>>6;
act = (vpx_variance8x8(y, y_stride, VP8_ZEROS, 0, &sse)+32)>>6;
#ifdef USE_SSD
sad = (vp8_variance8x8(y, y_stride, yd, yd_stride, &sse));
vpx_variance8x8(y, y_stride, yd, yd_stride, &sse);
sad = (sse + 32)>>6;
usad = (vp8_variance4x4(u, uv_stride, ud, uvd_stride, &sse));
vpx_variance4x4(u, uv_stride, ud, uvd_stride, &sse);
usad = (sse + 8)>>4;
vsad = (vp8_variance4x4(v, uv_stride, vd, uvd_stride, &sse));
vpx_variance4x4(v, uv_stride, vd, uvd_stride, &sse);
vsad = (sse + 8)>>4;
#else
sad = (vp8_sad8x8(y, y_stride, yd, yd_stride, UINT_MAX) + 32) >> 6;
usad = (vp8_sad4x4(u, uv_stride, ud, uvd_stride, UINT_MAX) + 8) >> 4;
vsad = (vp8_sad4x4(v, uv_stride, vd, uvd_stride, UINT_MAX) + 8) >> 4;
sad = (vpx_sad8x8(y, y_stride, yd, yd_stride) + 32) >> 6;
usad = (vpx_sad4x4(u, uv_stride, ud, uvd_stride) + 8) >> 4;
vsad = (vpx_sad4x4(v, uv_stride, vd, uvd_stride) + 8) >> 4;
#endif
}
@ -231,9 +232,9 @@ static void multiframe_quality_enhance_block
{
vp8_copy_mem8x8(y, y_stride, yd, yd_stride);
for (up = u, udp = ud, i = 0; i < uvblksize; ++i, up += uv_stride, udp += uvd_stride)
vpx_memcpy(udp, up, uvblksize);
memcpy(udp, up, uvblksize);
for (vp = v, vdp = vd, i = 0; i < uvblksize; ++i, vp += uv_stride, vdp += uvd_stride)
vpx_memcpy(vdp, vp, uvblksize);
memcpy(vdp, vp, uvblksize);
}
}
}
@ -341,8 +342,8 @@ void vp8_multiframe_quality_enhance
for (k = 0; k < 4; ++k, up += show->uv_stride, udp += dest->uv_stride,
vp += show->uv_stride, vdp += dest->uv_stride)
{
vpx_memcpy(udp, up, 4);
vpx_memcpy(vdp, vp, 4);
memcpy(udp, up, 4);
memcpy(vdp, vp, 4);
}
}
}

12
media/libvpx/vp8/common/postproc.c
Просмотреть файл

@ -355,8 +355,8 @@ void vp8_deblock(VP8_COMMON *cm,
else
mb_ppl = (unsigned char)ppl;
vpx_memset(ylptr, mb_ppl, 16);
vpx_memset(uvlptr, mb_ppl, 8);
memset(ylptr, mb_ppl, 16);
memset(uvlptr, mb_ppl, 8);
ylptr += 16;
uvlptr += 8;
@ -403,7 +403,7 @@ void vp8_de_noise(VP8_COMMON *cm,
(void) low_var_thresh;
(void) flag;
vpx_memset(limits, (unsigned char)ppl, 16 * mb_cols);
memset(limits, (unsigned char)ppl, 16 * mb_cols);
/* TODO: The original code don't filter the 2 outer rows and columns. */
for (mbr = 0; mbr < mb_rows; mbr++)
@ -427,7 +427,7 @@ void vp8_de_noise(VP8_COMMON *cm,
}
}
double vp8_gaussian(double sigma, double mu, double x)
static double gaussian(double sigma, double mu, double x)
{
return 1 / (sigma * sqrt(2.0 * 3.14159265)) *
(exp(-(x - mu) * (x - mu) / (2 * sigma * sigma)));
@ -455,7 +455,7 @@ static void fillrd(struct postproc_state *state, int q, int a)
for (i = -32; i < 32; i++)
{
const int v = (int)(.5 + 256 * vp8_gaussian(sigma, 0, i));
const int v = (int)(.5 + 256 * gaussian(sigma, 0, i));
if (v)
{
@ -763,7 +763,7 @@ int vp8_post_proc_frame(VP8_COMMON *oci, YV12_BUFFER_CONFIG *dest, vp8_ppflags_t
/* insure that postproc is set to all 0's so that post proc
* doesn't pull random data in from edge
*/
vpx_memset((&oci->post_proc_buffer_int)->buffer_alloc,128,(&oci->post_proc_buffer)->frame_size);
memset((&oci->post_proc_buffer_int)->buffer_alloc,128,(&oci->post_proc_buffer)->frame_size);
}
}

57
media/libvpx/vp8/common/reconinter.c
Просмотреть файл

@ -10,6 +10,8 @@
#include <limits.h>
#include <string.h>
#include "vpx_config.h"
#include "vp8_rtcd.h"
#include "vpx/vpx_integer.h"
@ -30,31 +32,8 @@ void vp8_copy_mem16x16_c(
for (r = 0; r < 16; r++)
{
#if !(CONFIG_FAST_UNALIGNED)
dst[0] = src[0];
dst[1] = src[1];
dst[2] = src[2];
dst[3] = src[3];
dst[4] = src[4];
dst[5] = src[5];
dst[6] = src[6];
dst[7] = src[7];
dst[8] = src[8];
dst[9] = src[9];
dst[10] = src[10];
dst[11] = src[11];
dst[12] = src[12];
dst[13] = src[13];
dst[14] = src[14];
dst[15] = src[15];
memcpy(dst, src, 16);
#else
((uint32_t *)dst)[0] = ((uint32_t *)src)[0] ;
((uint32_t *)dst)[1] = ((uint32_t *)src)[1] ;
((uint32_t *)dst)[2] = ((uint32_t *)src)[2] ;
((uint32_t *)dst)[3] = ((uint32_t *)src)[3] ;
#endif
src += src_stride;
dst += dst_stride;
@ -72,19 +51,8 @@ void vp8_copy_mem8x8_c(
for (r = 0; r < 8; r++)
{
#if !(CONFIG_FAST_UNALIGNED)
dst[0] = src[0];
dst[1] = src[1];
dst[2] = src[2];
dst[3] = src[3];
dst[4] = src[4];
dst[5] = src[5];
dst[6] = src[6];
dst[7] = src[7];
#else
((uint32_t *)dst)[0] = ((uint32_t *)src)[0] ;
((uint32_t *)dst)[1] = ((uint32_t *)src)[1] ;
#endif
memcpy(dst, src, 8);
src += src_stride;
dst += dst_stride;
@ -102,19 +70,8 @@ void vp8_copy_mem8x4_c(
for (r = 0; r < 4; r++)
{
#if !(CONFIG_FAST_UNALIGNED)
dst[0] = src[0];
dst[1] = src[1];
dst[2] = src[2];
dst[3] = src[3];
dst[4] = src[4];
dst[5] = src[5];
dst[6] = src[6];
dst[7] = src[7];
#else
((uint32_t *)dst)[0] = ((uint32_t *)src)[0] ;
((uint32_t *)dst)[1] = ((uint32_t *)src)[1] ;
#endif
memcpy(dst, src, 8);
src += src_stride;
dst += dst_stride;

22
media/libvpx/vp8/common/reconintra.c
Просмотреть файл

@ -70,10 +70,10 @@ void vp8_build_intra_predictors_mby_s_c(MACROBLOCKD *x,
expected_dc = 128;
}
/*vpx_memset(ypred_ptr, expected_dc, 256);*/
/*memset(ypred_ptr, expected_dc, 256);*/
for (r = 0; r < 16; r++)
{
vpx_memset(ypred_ptr, expected_dc, 16);
memset(ypred_ptr, expected_dc, 16);
ypred_ptr += y_stride;
}
}
@ -98,7 +98,7 @@ void vp8_build_intra_predictors_mby_s_c(MACROBLOCKD *x,
for (r = 0; r < 16; r++)
{
vpx_memset(ypred_ptr, yleft_col[r], 16);
memset(ypred_ptr, yleft_col[r], 16);
ypred_ptr += y_stride;
}
@ -202,12 +202,12 @@ void vp8_build_intra_predictors_mbuv_s_c(MACROBLOCKD *x,
}
/*vpx_memset(upred_ptr,expected_udc,64);*/
/*vpx_memset(vpred_ptr,expected_vdc,64);*/
/*memset(upred_ptr,expected_udc,64);*/
/*memset(vpred_ptr,expected_vdc,64);*/
for (i = 0; i < 8; i++)
{
vpx_memset(upred_ptr, expected_udc, 8);
vpx_memset(vpred_ptr, expected_vdc, 8);
memset(upred_ptr, expected_udc, 8);
memset(vpred_ptr, expected_vdc, 8);
upred_ptr += pred_stride;
vpred_ptr += pred_stride;
}
@ -217,8 +217,8 @@ void vp8_build_intra_predictors_mbuv_s_c(MACROBLOCKD *x,
{
for (i = 0; i < 8; i++)
{
vpx_memcpy(upred_ptr, uabove_row, 8);
vpx_memcpy(vpred_ptr, vabove_row, 8);
memcpy(upred_ptr, uabove_row, 8);
memcpy(vpred_ptr, vabove_row, 8);
upred_ptr += pred_stride;
vpred_ptr += pred_stride;
}
@ -229,8 +229,8 @@ void vp8_build_intra_predictors_mbuv_s_c(MACROBLOCKD *x,
{
for (i = 0; i < 8; i++)
{
vpx_memset(upred_ptr, uleft_col[i], 8);
vpx_memset(vpred_ptr, vleft_col[i], 8);
memset(upred_ptr, uleft_col[i], 8);
memset(vpred_ptr, vleft_col[i], 8);
upred_ptr += pred_stride;
vpred_ptr += pred_stride;
}

6
media/libvpx/vp8/common/rtcd.c
Просмотреть файл

@ -7,15 +7,13 @@
* in the file PATENTS. All contributing project authors may
* be found in the AUTHORS file in the root of the source tree.
*/
#include "vpx_config.h"
#include "./vpx_config.h"
#define RTCD_C
#include "vp8_rtcd.h"
#include "./vp8_rtcd.h"
#include "vpx_ports/vpx_once.h"
extern void vpx_scale_rtcd(void);
void vp8_rtcd()
{
vpx_scale_rtcd();
once(setup_rtcd_internal);
}

302
media/libvpx/vp8/common/sad_c.c
Просмотреть файл

@ -1,302 +0,0 @@
/*
* Copyright (c) 2010 The WebM project authors. All Rights Reserved.
*
* Use of this source code is governed by a BSD-style license
* that can be found in the LICENSE file in the root of the source
* tree. An additional intellectual property rights grant can be found
* in the file PATENTS. All contributing project authors may
* be found in the AUTHORS file in the root of the source tree.
*/
#include <limits.h>
#include <stdlib.h>
#include "vpx_config.h"
#include "vpx/vpx_integer.h"
static unsigned int sad_mx_n_c(const unsigned char *src_ptr, int src_stride,
const unsigned char *ref_ptr, int ref_stride,
unsigned int max_sad, int m, int n)
{
int r, c;
unsigned int sad = 0;
for (r = 0; r < n; r++)
{
for (c = 0; c < m; c++)
{
sad += abs(src_ptr[c] - ref_ptr[c]);
}
if (sad > max_sad)
break;
src_ptr += src_stride;
ref_ptr += ref_stride;
}
return sad;
}
/* max_sad is provided as an optional optimization point. Alternative
* implementations of these functions are not required to check it.
*/
unsigned int vp8_sad16x16_c(const unsigned char *src_ptr, int src_stride,
const unsigned char *ref_ptr, int ref_stride,
unsigned int max_sad)
{
return sad_mx_n_c(src_ptr, src_stride, ref_ptr, ref_stride, max_sad, 16, 16);
}
unsigned int vp8_sad8x8_c(const unsigned char *src_ptr, int src_stride,
const unsigned char *ref_ptr, int ref_stride,
unsigned int max_sad)
{
return sad_mx_n_c(src_ptr, src_stride, ref_ptr, ref_stride, max_sad, 8, 8);
}
unsigned int vp8_sad16x8_c(const unsigned char *src_ptr, int src_stride,
const unsigned char *ref_ptr, int ref_stride,
unsigned int max_sad)
{
return sad_mx_n_c(src_ptr, src_stride, ref_ptr, ref_stride, max_sad, 16, 8);
}
unsigned int vp8_sad8x16_c(const unsigned char *src_ptr, int src_stride,
const unsigned char *ref_ptr, int ref_stride,
unsigned int max_sad)
{
return sad_mx_n_c(src_ptr, src_stride, ref_ptr, ref_stride, max_sad, 8, 16);
}
unsigned int vp8_sad4x4_c(const unsigned char *src_ptr, int src_stride,
const unsigned char *ref_ptr, int ref_stride,
unsigned int max_sad)
{
return sad_mx_n_c(src_ptr, src_stride, ref_ptr, ref_stride, max_sad, 4, 4);
}
void vp8_sad16x16x3_c(const unsigned char *src_ptr, int src_stride,
const unsigned char *ref_ptr, int ref_stride,
unsigned int *sad_array)
{
sad_array[0] = vp8_sad16x16_c(src_ptr, src_stride, ref_ptr + 0, ref_stride, UINT_MAX);
sad_array[1] = vp8_sad16x16_c(src_ptr, src_stride, ref_ptr + 1, ref_stride, UINT_MAX);
sad_array[2] = vp8_sad16x16_c(src_ptr, src_stride, ref_ptr + 2, ref_stride, UINT_MAX);
}
void vp8_sad16x16x8_c(const unsigned char *src_ptr, int src_stride,
const unsigned char *ref_ptr, int ref_stride,
unsigned short *sad_array)
{
sad_array[0] = (unsigned short)vp8_sad16x16_c(src_ptr, src_stride, ref_ptr + 0, ref_stride, UINT_MAX);
sad_array[1] = (unsigned short)vp8_sad16x16_c(src_ptr, src_stride, ref_ptr + 1, ref_stride, UINT_MAX);
sad_array[2] = (unsigned short)vp8_sad16x16_c(src_ptr, src_stride, ref_ptr + 2, ref_stride, UINT_MAX);
sad_array[3] = (unsigned short)vp8_sad16x16_c(src_ptr, src_stride, ref_ptr + 3, ref_stride, UINT_MAX);
sad_array[4] = (unsigned short)vp8_sad16x16_c(src_ptr, src_stride, ref_ptr + 4, ref_stride, UINT_MAX);
sad_array[5] = (unsigned short)vp8_sad16x16_c(src_ptr, src_stride, ref_ptr + 5, ref_stride, UINT_MAX);
sad_array[6] = (unsigned short)vp8_sad16x16_c(src_ptr, src_stride, ref_ptr + 6, ref_stride, UINT_MAX);
sad_array[7] = (unsigned short)vp8_sad16x16_c(src_ptr, src_stride, ref_ptr + 7, ref_stride, UINT_MAX);
}
void vp8_sad16x8x3_c(const unsigned char *src_ptr, int src_stride,
const unsigned char *ref_ptr, int ref_stride,
unsigned int *sad_array)
{
sad_array[0] = vp8_sad16x8_c(src_ptr, src_stride, ref_ptr + 0, ref_stride, UINT_MAX);
sad_array[1] = vp8_sad16x8_c(src_ptr, src_stride, ref_ptr + 1, ref_stride, UINT_MAX);
sad_array[2] = vp8_sad16x8_c(src_ptr, src_stride, ref_ptr + 2, ref_stride, UINT_MAX);
}
void vp8_sad16x8x8_c(const unsigned char *src_ptr, int src_stride,
const unsigned char *ref_ptr, int ref_stride,
unsigned short *sad_array)
{
sad_array[0] = (unsigned short)vp8_sad16x8_c(src_ptr, src_stride, ref_ptr + 0, ref_stride, UINT_MAX);
sad_array[1] = (unsigned short)vp8_sad16x8_c(src_ptr, src_stride, ref_ptr + 1, ref_stride, UINT_MAX);
sad_array[2] = (unsigned short)vp8_sad16x8_c(src_ptr, src_stride, ref_ptr + 2, ref_stride, UINT_MAX);
sad_array[3] = (unsigned short)vp8_sad16x8_c(src_ptr, src_stride, ref_ptr + 3, ref_stride, UINT_MAX);
sad_array[4] = (unsigned short)vp8_sad16x8_c(src_ptr, src_stride, ref_ptr + 4, ref_stride, UINT_MAX);
sad_array[5] = (unsigned short)vp8_sad16x8_c(src_ptr, src_stride, ref_ptr + 5, ref_stride, UINT_MAX);
sad_array[6] = (unsigned short)vp8_sad16x8_c(src_ptr, src_stride, ref_ptr + 6, ref_stride, UINT_MAX);
sad_array[7] = (unsigned short)vp8_sad16x8_c(src_ptr, src_stride, ref_ptr + 7, ref_stride, UINT_MAX);
}
void vp8_sad8x8x3_c(const unsigned char *src_ptr, int src_stride,
const unsigned char *ref_ptr, int ref_stride,
unsigned int *sad_array)
{
sad_array[0] = vp8_sad8x8_c(src_ptr, src_stride, ref_ptr + 0, ref_stride, UINT_MAX);
sad_array[1] = vp8_sad8x8_c(src_ptr, src_stride, ref_ptr + 1, ref_stride, UINT_MAX);
sad_array[2] = vp8_sad8x8_c(src_ptr, src_stride, ref_ptr + 2, ref_stride, UINT_MAX);
}
void vp8_sad8x8x8_c(const unsigned char *src_ptr, int src_stride,
const unsigned char *ref_ptr, int ref_stride,
unsigned short *sad_array)
{
sad_array[0] = (unsigned short)vp8_sad8x8_c(src_ptr, src_stride, ref_ptr + 0, ref_stride, UINT_MAX);
sad_array[1] = (unsigned short)vp8_sad8x8_c(src_ptr, src_stride, ref_ptr + 1, ref_stride, UINT_MAX);
sad_array[2] = (unsigned short)vp8_sad8x8_c(src_ptr, src_stride, ref_ptr + 2, ref_stride, UINT_MAX);
sad_array[3] = (unsigned short)vp8_sad8x8_c(src_ptr, src_stride, ref_ptr + 3, ref_stride, UINT_MAX);
sad_array[4] = (unsigned short)vp8_sad8x8_c(src_ptr, src_stride, ref_ptr + 4, ref_stride, UINT_MAX);
sad_array[5] = (unsigned short)vp8_sad8x8_c(src_ptr, src_stride, ref_ptr + 5, ref_stride, UINT_MAX);
sad_array[6] = (unsigned short)vp8_sad8x8_c(src_ptr, src_stride, ref_ptr + 6, ref_stride, UINT_MAX);
sad_array[7] = (unsigned short)vp8_sad8x8_c(src_ptr, src_stride, ref_ptr + 7, ref_stride, UINT_MAX);
}
void vp8_sad8x16x3_c(const unsigned char *src_ptr, int src_stride,
const unsigned char *ref_ptr, int ref_stride,
unsigned int *sad_array)
{
sad_array[0] = vp8_sad8x16_c(src_ptr, src_stride, ref_ptr + 0, ref_stride, UINT_MAX);
sad_array[1] = vp8_sad8x16_c(src_ptr, src_stride, ref_ptr + 1, ref_stride, UINT_MAX);
sad_array[2] = vp8_sad8x16_c(src_ptr, src_stride, ref_ptr + 2, ref_stride, UINT_MAX);
}
void vp8_sad8x16x8_c(const unsigned char *src_ptr, int src_stride,
const unsigned char *ref_ptr, int ref_stride,
unsigned short *sad_array)
{
sad_array[0] = (unsigned short)vp8_sad8x16_c(src_ptr, src_stride, ref_ptr + 0, ref_stride, UINT_MAX);
sad_array[1] = (unsigned short)vp8_sad8x16_c(src_ptr, src_stride, ref_ptr + 1, ref_stride, UINT_MAX);
sad_array[2] = (unsigned short)vp8_sad8x16_c(src_ptr, src_stride, ref_ptr + 2, ref_stride, UINT_MAX);
sad_array[3] = (unsigned short)vp8_sad8x16_c(src_ptr, src_stride, ref_ptr + 3, ref_stride, UINT_MAX);
sad_array[4] = (unsigned short)vp8_sad8x16_c(src_ptr, src_stride, ref_ptr + 4, ref_stride, UINT_MAX);
sad_array[5] = (unsigned short)vp8_sad8x16_c(src_ptr, src_stride, ref_ptr + 5, ref_stride, UINT_MAX);
sad_array[6] = (unsigned short)vp8_sad8x16_c(src_ptr, src_stride, ref_ptr + 6, ref_stride, UINT_MAX);
sad_array[7] = (unsigned short)vp8_sad8x16_c(src_ptr, src_stride, ref_ptr + 7, ref_stride, UINT_MAX);
}
void vp8_sad4x4x3_c(const unsigned char *src_ptr, int src_stride,
const unsigned char *ref_ptr, int ref_stride,
unsigned int *sad_array)
{
sad_array[0] = vp8_sad4x4_c(src_ptr, src_stride, ref_ptr + 0, ref_stride, UINT_MAX);
sad_array[1] = vp8_sad4x4_c(src_ptr, src_stride, ref_ptr + 1, ref_stride, UINT_MAX);
sad_array[2] = vp8_sad4x4_c(src_ptr, src_stride, ref_ptr + 2, ref_stride, UINT_MAX);
}
void vp8_sad4x4x8_c(const unsigned char *src_ptr, int src_stride,
const unsigned char *ref_ptr, int ref_stride,
unsigned short *sad_array)
{
sad_array[0] = (unsigned short)vp8_sad4x4_c(src_ptr, src_stride, ref_ptr + 0, ref_stride, UINT_MAX);
sad_array[1] = (unsigned short)vp8_sad4x4_c(src_ptr, src_stride, ref_ptr + 1, ref_stride, UINT_MAX);
sad_array[2] = (unsigned short)vp8_sad4x4_c(src_ptr, src_stride, ref_ptr + 2, ref_stride, UINT_MAX);
sad_array[3] = (unsigned short)vp8_sad4x4_c(src_ptr, src_stride, ref_ptr + 3, ref_stride, UINT_MAX);
sad_array[4] = (unsigned short)vp8_sad4x4_c(src_ptr, src_stride, ref_ptr + 4, ref_stride, UINT_MAX);
sad_array[5] = (unsigned short)vp8_sad4x4_c(src_ptr, src_stride, ref_ptr + 5, ref_stride, UINT_MAX);
sad_array[6] = (unsigned short)vp8_sad4x4_c(src_ptr, src_stride, ref_ptr + 6, ref_stride, UINT_MAX);
sad_array[7] = (unsigned short)vp8_sad4x4_c(src_ptr, src_stride, ref_ptr + 7, ref_stride, UINT_MAX);
}
void vp8_sad16x16x4d_c(const unsigned char *src_ptr, int src_stride,
const unsigned char * const ref_ptr[], int ref_stride,
unsigned int *sad_array)
{
sad_array[0] = vp8_sad16x16_c(src_ptr, src_stride, ref_ptr[0], ref_stride, UINT_MAX);
sad_array[1] = vp8_sad16x16_c(src_ptr, src_stride, ref_ptr[1], ref_stride, UINT_MAX);
sad_array[2] = vp8_sad16x16_c(src_ptr, src_stride, ref_ptr[2], ref_stride, UINT_MAX);
sad_array[3] = vp8_sad16x16_c(src_ptr, src_stride, ref_ptr[3], ref_stride, UINT_MAX);
}
void vp8_sad16x8x4d_c(const unsigned char *src_ptr, int src_stride,
const unsigned char * const ref_ptr[], int ref_stride,
unsigned int *sad_array)
{
sad_array[0] = vp8_sad16x8_c(src_ptr, src_stride, ref_ptr[0], ref_stride, UINT_MAX);
sad_array[1] = vp8_sad16x8_c(src_ptr, src_stride, ref_ptr[1], ref_stride, UINT_MAX);
sad_array[2] = vp8_sad16x8_c(src_ptr, src_stride, ref_ptr[2], ref_stride, UINT_MAX);
sad_array[3] = vp8_sad16x8_c(src_ptr, src_stride, ref_ptr[3], ref_stride, UINT_MAX);
}
void vp8_sad8x8x4d_c(const unsigned char *src_ptr, int src_stride,
const unsigned char * const ref_ptr[], int ref_stride,
unsigned int *sad_array)
{
sad_array[0] = vp8_sad8x8_c(src_ptr, src_stride, ref_ptr[0], ref_stride, UINT_MAX);
sad_array[1] = vp8_sad8x8_c(src_ptr, src_stride, ref_ptr[1], ref_stride, UINT_MAX);
sad_array[2] = vp8_sad8x8_c(src_ptr, src_stride, ref_ptr[2], ref_stride, UINT_MAX);
sad_array[3] = vp8_sad8x8_c(src_ptr, src_stride, ref_ptr[3], ref_stride, UINT_MAX);
}
void vp8_sad8x16x4d_c(const unsigned char *src_ptr, int src_stride,
const unsigned char * const ref_ptr[], int ref_stride,
unsigned int *sad_array)
{
sad_array[0] = vp8_sad8x16_c(src_ptr, src_stride, ref_ptr[0], ref_stride, UINT_MAX);
sad_array[1] = vp8_sad8x16_c(src_ptr, src_stride, ref_ptr[1], ref_stride, UINT_MAX);
sad_array[2] = vp8_sad8x16_c(src_ptr, src_stride, ref_ptr[2], ref_stride, UINT_MAX);
sad_array[3] = vp8_sad8x16_c(src_ptr, src_stride, ref_ptr[3], ref_stride, UINT_MAX);
}
void vp8_sad4x4x4d_c(const unsigned char *src_ptr, int src_stride,
const unsigned char * const ref_ptr[], int ref_stride,
unsigned int *sad_array)
{
sad_array[0] = vp8_sad4x4_c(src_ptr, src_stride, ref_ptr[0], ref_stride, UINT_MAX);
sad_array[1] = vp8_sad4x4_c(src_ptr, src_stride, ref_ptr[1], ref_stride, UINT_MAX);
sad_array[2] = vp8_sad4x4_c(src_ptr, src_stride, ref_ptr[2], ref_stride, UINT_MAX);
sad_array[3] = vp8_sad4x4_c(src_ptr, src_stride, ref_ptr[3], ref_stride, UINT_MAX);
}
/* Copy 2 macroblocks to a buffer */
void vp8_copy32xn_c(unsigned char *src_ptr, int src_stride,
unsigned char *dst_ptr, int dst_stride,
int height)
{
int r;
for (r = 0; r < height; r++)
{
#if !(CONFIG_FAST_UNALIGNED)
dst_ptr[0] = src_ptr[0];
dst_ptr[1] = src_ptr[1];
dst_ptr[2] = src_ptr[2];
dst_ptr[3] = src_ptr[3];
dst_ptr[4] = src_ptr[4];
dst_ptr[5] = src_ptr[5];
dst_ptr[6] = src_ptr[6];
dst_ptr[7] = src_ptr[7];
dst_ptr[8] = src_ptr[8];
dst_ptr[9] = src_ptr[9];
dst_ptr[10] = src_ptr[10];
dst_ptr[11] = src_ptr[11];
dst_ptr[12] = src_ptr[12];
dst_ptr[13] = src_ptr[13];
dst_ptr[14] = src_ptr[14];
dst_ptr[15] = src_ptr[15];
dst_ptr[16] = src_ptr[16];
dst_ptr[17] = src_ptr[17];
dst_ptr[18] = src_ptr[18];
dst_ptr[19] = src_ptr[19];
dst_ptr[20] = src_ptr[20];
dst_ptr[21] = src_ptr[21];
dst_ptr[22] = src_ptr[22];
dst_ptr[23] = src_ptr[23];
dst_ptr[24] = src_ptr[24];
dst_ptr[25] = src_ptr[25];
dst_ptr[26] = src_ptr[26];
dst_ptr[27] = src_ptr[27];
dst_ptr[28] = src_ptr[28];
dst_ptr[29] = src_ptr[29];
dst_ptr[30] = src_ptr[30];
dst_ptr[31] = src_ptr[31];
#else
((uint32_t *)dst_ptr)[0] = ((uint32_t *)src_ptr)[0] ;
((uint32_t *)dst_ptr)[1] = ((uint32_t *)src_ptr)[1] ;
((uint32_t *)dst_ptr)[2] = ((uint32_t *)src_ptr)[2] ;
((uint32_t *)dst_ptr)[3] = ((uint32_t *)src_ptr)[3] ;
((uint32_t *)dst_ptr)[4] = ((uint32_t *)src_ptr)[4] ;
((uint32_t *)dst_ptr)[5] = ((uint32_t *)src_ptr)[5] ;
((uint32_t *)dst_ptr)[6] = ((uint32_t *)src_ptr)[6] ;
((uint32_t *)dst_ptr)[7] = ((uint32_t *)src_ptr)[7] ;
#endif
src_ptr += src_stride;
dst_ptr += dst_stride;
}
}

12
media/libvpx/vp8/common/setupintrarecon.c
Просмотреть файл

@ -17,15 +17,15 @@ void vp8_setup_intra_recon(YV12_BUFFER_CONFIG *ybf)
int i;
/* set up frame new frame for intra coded blocks */
vpx_memset(ybf->y_buffer - 1 - ybf->y_stride, 127, ybf->y_width + 5);
memset(ybf->y_buffer - 1 - ybf->y_stride, 127, ybf->y_width + 5);
for (i = 0; i < ybf->y_height; i++)
ybf->y_buffer[ybf->y_stride *i - 1] = (unsigned char) 129;
vpx_memset(ybf->u_buffer - 1 - ybf->uv_stride, 127, ybf->uv_width + 5);
memset(ybf->u_buffer - 1 - ybf->uv_stride, 127, ybf->uv_width + 5);
for (i = 0; i < ybf->uv_height; i++)
ybf->u_buffer[ybf->uv_stride *i - 1] = (unsigned char) 129;
vpx_memset(ybf->v_buffer - 1 - ybf->uv_stride, 127, ybf->uv_width + 5);
memset(ybf->v_buffer - 1 - ybf->uv_stride, 127, ybf->uv_width + 5);
for (i = 0; i < ybf->uv_height; i++)
ybf->v_buffer[ybf->uv_stride *i - 1] = (unsigned char) 129;
@ -33,7 +33,7 @@ void vp8_setup_intra_recon(YV12_BUFFER_CONFIG *ybf)
void vp8_setup_intra_recon_top_line(YV12_BUFFER_CONFIG *ybf)
{
vpx_memset(ybf->y_buffer - 1 - ybf->y_stride, 127, ybf->y_width + 5);
vpx_memset(ybf->u_buffer - 1 - ybf->uv_stride, 127, ybf->uv_width + 5);
vpx_memset(ybf->v_buffer - 1 - ybf->uv_stride, 127, ybf->uv_width + 5);
memset(ybf->y_buffer - 1 - ybf->y_stride, 127, ybf->y_width + 5);
memset(ybf->u_buffer - 1 - ybf->uv_stride, 127, ybf->uv_width + 5);
memset(ybf->v_buffer - 1 - ybf->uv_stride, 127, ybf->uv_width + 5);
}

71
media/libvpx/vp8/common/variance.h
Просмотреть файл

@ -14,50 +14,42 @@
#include "vpx_config.h"
#include "vpx/vpx_integer.h"
#ifdef __cplusplus
extern "C" {
#endif
typedef unsigned int(*vp8_sad_fn_t)(
const unsigned char *src_ptr,
typedef unsigned int(*vpx_sad_fn_t)(
const uint8_t *src_ptr,
int source_stride,
const unsigned char *ref_ptr,
int ref_stride,
unsigned int max_sad);
const uint8_t *ref_ptr,
int ref_stride);
typedef void (*vp8_copy32xn_fn_t)(
const unsigned char *src_ptr,
int source_stride,
const unsigned char *ref_ptr,
unsigned char *ref_ptr,
int ref_stride,
int n);
typedef void (*vp8_sad_multi_fn_t)(
typedef void (*vpx_sad_multi_fn_t)(
const unsigned char *src_ptr,
int source_stride,
const unsigned char *ref_ptr,
const unsigned char *ref_array,
int ref_stride,
unsigned int *sad_array);
typedef void (*vp8_sad_multi1_fn_t)
typedef void (*vpx_sad_multi_d_fn_t)
(
const unsigned char *src_ptr,
int source_stride,
const unsigned char *ref_ptr,
int ref_stride,
unsigned short *sad_array
);
typedef void (*vp8_sad_multi_d_fn_t)
(
const unsigned char *src_ptr,
int source_stride,
const unsigned char * const ref_ptr[],
const unsigned char * const ref_array[],
int ref_stride,
unsigned int *sad_array
);
typedef unsigned int (*vp8_variance_fn_t)
typedef unsigned int (*vpx_variance_fn_t)
(
const unsigned char *src_ptr,
int source_stride,
@ -77,40 +69,17 @@ typedef unsigned int (*vp8_subpixvariance_fn_t)
unsigned int *sse
);
typedef void (*vp8_ssimpf_fn_t)
(
unsigned char *s,
int sp,
unsigned char *r,
int rp,
unsigned long *sum_s,
unsigned long *sum_r,
unsigned long *sum_sq_s,
unsigned long *sum_sq_r,
unsigned long *sum_sxr
);
typedef unsigned int (*vp8_getmbss_fn_t)(const short *);
typedef unsigned int (*vp8_get16x16prederror_fn_t)
(
const unsigned char *src_ptr,
int source_stride,
const unsigned char *ref_ptr,
int ref_stride
);
typedef struct variance_vtable
{
vp8_sad_fn_t sdf;
vp8_variance_fn_t vf;
vpx_sad_fn_t sdf;
vpx_variance_fn_t vf;
vp8_subpixvariance_fn_t svf;
vp8_variance_fn_t svf_halfpix_h;
vp8_variance_fn_t svf_halfpix_v;
vp8_variance_fn_t svf_halfpix_hv;
vp8_sad_multi_fn_t sdx3f;
vp8_sad_multi1_fn_t sdx8f;
vp8_sad_multi_d_fn_t sdx4df;
vpx_variance_fn_t svf_halfpix_h;
vpx_variance_fn_t svf_halfpix_v;
vpx_variance_fn_t svf_halfpix_hv;
vpx_sad_multi_fn_t sdx3f;
vpx_sad_multi_fn_t sdx8f;
vpx_sad_multi_d_fn_t sdx4df;
#if ARCH_X86 || ARCH_X86_64
vp8_copy32xn_fn_t copymem;
#endif

167
media/libvpx/vp8/common/variance_c.c
Просмотреть файл

@ -8,43 +8,34 @@
* be found in the AUTHORS file in the root of the source tree.
*/
#include "variance.h"
#include "./vp8_rtcd.h"
#include "filter.h"
#include "variance.h"
unsigned int vp8_get_mb_ss_c
(
const short *src_ptr
)
{
unsigned int i = 0, sum = 0;
do
{
sum += (src_ptr[i] * src_ptr[i]);
i++;
}
while (i < 256);
return sum;
/* This is a bad idea.
* ctz = count trailing zeros */
static int ctz(int a) {
int b = 0;
while (a != 1) {
a >>= 1;
b++;
}
return b;
}
static void variance(
static unsigned int variance(
const unsigned char *src_ptr,
int source_stride,
const unsigned char *ref_ptr,
int recon_stride,
int w,
int h,
unsigned int *sse,
int *sum)
unsigned int *sse)
{
int i, j;
int diff;
int diff, sum;
*sum = 0;
sum = 0;
*sse = 0;
for (i = 0; i < h; i++)
@ -52,114 +43,17 @@ static void variance(
for (j = 0; j < w; j++)
{
diff = src_ptr[j] - ref_ptr[j];
*sum += diff;
sum += diff;
*sse += diff * diff;
}
src_ptr += source_stride;
ref_ptr += recon_stride;
}
return (*sse - (((unsigned int)sum * sum) >> (int)((ctz(w) + ctz(h)))));
}
unsigned int vp8_variance16x16_c(
const unsigned char *src_ptr,
int source_stride,
const unsigned char *ref_ptr,
int recon_stride,
unsigned int *sse)
{
unsigned int var;
int avg;
variance(src_ptr, source_stride, ref_ptr, recon_stride, 16, 16, &var, &avg);
*sse = var;
return (var - (((unsigned int)avg * avg) >> 8));
}
unsigned int vp8_variance8x16_c(
const unsigned char *src_ptr,
int source_stride,
const unsigned char *ref_ptr,
int recon_stride,
unsigned int *sse)
{
unsigned int var;
int avg;
variance(src_ptr, source_stride, ref_ptr, recon_stride, 8, 16, &var, &avg);
*sse = var;
return (var - (((unsigned int)avg * avg) >> 7));
}
unsigned int vp8_variance16x8_c(
const unsigned char *src_ptr,
int source_stride,
const unsigned char *ref_ptr,
int recon_stride,
unsigned int *sse)
{
unsigned int var;
int avg;
variance(src_ptr, source_stride, ref_ptr, recon_stride, 16, 8, &var, &avg);
*sse = var;
return (var - (((unsigned int)avg * avg) >> 7));
}
unsigned int vp8_variance8x8_c(
const unsigned char *src_ptr,
int source_stride,
const unsigned char *ref_ptr,
int recon_stride,
unsigned int *sse)
{
unsigned int var;
int avg;
variance(src_ptr, source_stride, ref_ptr, recon_stride, 8, 8, &var, &avg);
*sse = var;
return (var - (((unsigned int)avg * avg) >> 6));
}
unsigned int vp8_variance4x4_c(
const unsigned char *src_ptr,
int source_stride,
const unsigned char *ref_ptr,
int recon_stride,
unsigned int *sse)
{
unsigned int var;
int avg;
variance(src_ptr, source_stride, ref_ptr, recon_stride, 4, 4, &var, &avg);
*sse = var;
return (var - (((unsigned int)avg * avg) >> 4));
}
unsigned int vp8_mse16x16_c(
const unsigned char *src_ptr,
int source_stride,
const unsigned char *ref_ptr,
int recon_stride,
unsigned int *sse)
{
unsigned int var;
int avg;
variance(src_ptr, source_stride, ref_ptr, recon_stride, 16, 16, &var, &avg);
*sse = var;
return var;
}
/****************************************************************************
*
* ROUTINE : filter_block2d_bil_first_pass
@ -303,7 +197,7 @@ unsigned int vp8_sub_pixel_variance4x4_c
/* Now filter Verticaly */
var_filter_block2d_bil_second_pass(FData3, temp2, 4, 4, 4, 4, VFilter);
return vp8_variance4x4_c(temp2, 4, dst_ptr, dst_pixels_per_line, sse);
return variance(temp2, 4, dst_ptr, dst_pixels_per_line, 4, 4, sse);
}
@ -328,7 +222,7 @@ unsigned int vp8_sub_pixel_variance8x8_c
var_filter_block2d_bil_first_pass(src_ptr, FData3, src_pixels_per_line, 1, 9, 8, HFilter);
var_filter_block2d_bil_second_pass(FData3, temp2, 8, 8, 8, 8, VFilter);
return vp8_variance8x8_c(temp2, 8, dst_ptr, dst_pixels_per_line, sse);
return variance(temp2, 8, dst_ptr, dst_pixels_per_line, 8, 8, sse);
}
unsigned int vp8_sub_pixel_variance16x16_c
@ -352,7 +246,7 @@ unsigned int vp8_sub_pixel_variance16x16_c
var_filter_block2d_bil_first_pass(src_ptr, FData3, src_pixels_per_line, 1, 17, 16, HFilter);
var_filter_block2d_bil_second_pass(FData3, temp2, 16, 16, 16, 16, VFilter);
return vp8_variance16x16_c(temp2, 16, dst_ptr, dst_pixels_per_line, sse);
return variance(temp2, 16, dst_ptr, dst_pixels_per_line, 16, 16, sse);
}
@ -392,21 +286,6 @@ unsigned int vp8_variance_halfpixvar16x16_hv_c(
}
unsigned int vp8_sub_pixel_mse16x16_c
(
const unsigned char *src_ptr,
int src_pixels_per_line,
int xoffset,
int yoffset,
const unsigned char *dst_ptr,
int dst_pixels_per_line,
unsigned int *sse
)
{
vp8_sub_pixel_variance16x16_c(src_ptr, src_pixels_per_line, xoffset, yoffset, dst_ptr, dst_pixels_per_line, sse);
return *sse;
}
unsigned int vp8_sub_pixel_variance16x8_c
(
const unsigned char *src_ptr,
@ -428,7 +307,7 @@ unsigned int vp8_sub_pixel_variance16x8_c
var_filter_block2d_bil_first_pass(src_ptr, FData3, src_pixels_per_line, 1, 9, 16, HFilter);
var_filter_block2d_bil_second_pass(FData3, temp2, 16, 16, 8, 16, VFilter);
return vp8_variance16x8_c(temp2, 16, dst_ptr, dst_pixels_per_line, sse);
return variance(temp2, 16, dst_ptr, dst_pixels_per_line, 16, 8, sse);
}
unsigned int vp8_sub_pixel_variance8x16_c
@ -454,5 +333,5 @@ unsigned int vp8_sub_pixel_variance8x16_c
var_filter_block2d_bil_first_pass(src_ptr, FData3, src_pixels_per_line, 1, 17, 8, HFilter);
var_filter_block2d_bil_second_pass(FData3, temp2, 8, 8, 16, 8, VFilter);
return vp8_variance8x16_c(temp2, 8, dst_ptr, dst_pixels_per_line, sse);
return variance(temp2, 8, dst_ptr, dst_pixels_per_line, 8, 16, sse);
}

93
media/libvpx/vp8/common/x86/copy_sse2.asm Normal file
Просмотреть файл

@ -0,0 +1,93 @@
;
; Copyright (c) 2010 The WebM project authors. All Rights Reserved.
;
; Use of this source code is governed by a BSD-style license
; that can be found in the LICENSE file in the root of the source
; tree. An additional intellectual property rights grant can be found
; in the file PATENTS. All contributing project authors may
; be found in the AUTHORS file in the root of the source tree.
;
%include "vpx_ports/x86_abi_support.asm"
;void vp8_copy32xn_sse2(
; unsigned char *src_ptr,
; int src_stride,
; unsigned char *dst_ptr,
; int dst_stride,
; int height);
global sym(vp8_copy32xn_sse2) PRIVATE
sym(vp8_copy32xn_sse2):
push rbp
mov rbp, rsp
SHADOW_ARGS_TO_STACK 5
SAVE_XMM 7
push rsi
push rdi
; end prolog
mov rsi, arg(0) ;src_ptr
mov rdi, arg(2) ;dst_ptr
movsxd rax, dword ptr arg(1) ;src_stride
movsxd rdx, dword ptr arg(3) ;dst_stride
movsxd rcx, dword ptr arg(4) ;height
.block_copy_sse2_loopx4:
movdqu xmm0, XMMWORD PTR [rsi]
movdqu xmm1, XMMWORD PTR [rsi + 16]
movdqu xmm2, XMMWORD PTR [rsi + rax]
movdqu xmm3, XMMWORD PTR [rsi + rax + 16]
lea rsi, [rsi+rax*2]
movdqu xmm4, XMMWORD PTR [rsi]
movdqu xmm5, XMMWORD PTR [rsi + 16]
movdqu xmm6, XMMWORD PTR [rsi + rax]
movdqu xmm7, XMMWORD PTR [rsi + rax + 16]
lea rsi, [rsi+rax*2]
movdqa XMMWORD PTR [rdi], xmm0
movdqa XMMWORD PTR [rdi + 16], xmm1
movdqa XMMWORD PTR [rdi + rdx], xmm2
movdqa XMMWORD PTR [rdi + rdx + 16], xmm3
lea rdi, [rdi+rdx*2]
movdqa XMMWORD PTR [rdi], xmm4
movdqa XMMWORD PTR [rdi + 16], xmm5
movdqa XMMWORD PTR [rdi + rdx], xmm6
movdqa XMMWORD PTR [rdi + rdx + 16], xmm7
lea rdi, [rdi+rdx*2]
sub rcx, 4
cmp rcx, 4
jge .block_copy_sse2_loopx4
cmp rcx, 0
je .copy_is_done
.block_copy_sse2_loop:
movdqu xmm0, XMMWORD PTR [rsi]
movdqu xmm1, XMMWORD PTR [rsi + 16]
lea rsi, [rsi+rax]
movdqa XMMWORD PTR [rdi], xmm0
movdqa XMMWORD PTR [rdi + 16], xmm1
lea rdi, [rdi+rdx]
sub rcx, 1
jne .block_copy_sse2_loop
.copy_is_done:
; begin epilog
pop rdi
pop rsi
RESTORE_XMM
UNSHADOW_ARGS
pop rbp
ret

146
media/libvpx/vp8/common/x86/copy_sse3.asm Normal file
Просмотреть файл

@ -0,0 +1,146 @@
;
; Copyright (c) 2010 The WebM project authors. All Rights Reserved.
;
; Use of this source code is governed by a BSD-style license
; that can be found in the LICENSE file in the root of the source
; tree. An additional intellectual property rights grant can be found
; in the file PATENTS. All contributing project authors may
; be found in the AUTHORS file in the root of the source tree.
;
%include "vpx_ports/x86_abi_support.asm"
%macro STACK_FRAME_CREATE_X3 0
%if ABI_IS_32BIT
%define src_ptr rsi
%define src_stride rax
%define ref_ptr rdi
%define ref_stride rdx
%define end_ptr rcx
%define ret_var rbx
%define result_ptr arg(4)
%define max_sad arg(4)
%define height dword ptr arg(4)
push rbp
mov rbp, rsp
push rsi
push rdi
push rbx
mov rsi, arg(0) ; src_ptr
mov rdi, arg(2) ; ref_ptr
movsxd rax, dword ptr arg(1) ; src_stride
movsxd rdx, dword ptr arg(3) ; ref_stride
%else
%if LIBVPX_YASM_WIN64
SAVE_XMM 7, u
%define src_ptr rcx
%define src_stride rdx
%define ref_ptr r8
%define ref_stride r9
%define end_ptr r10
%define ret_var r11
%define result_ptr [rsp+xmm_stack_space+8+4*8]
%define max_sad [rsp+xmm_stack_space+8+4*8]
%define height dword ptr [rsp+xmm_stack_space+8+4*8]
%else
%define src_ptr rdi
%define src_stride rsi
%define ref_ptr rdx
%define ref_stride rcx
%define end_ptr r9
%define ret_var r10
%define result_ptr r8
%define max_sad r8
%define height r8
%endif
%endif
%endmacro
%macro STACK_FRAME_DESTROY_X3 0
%define src_ptr
%define src_stride
%define ref_ptr
%define ref_stride
%define end_ptr
%define ret_var
%define result_ptr
%define max_sad
%define height
%if ABI_IS_32BIT
pop rbx
pop rdi
pop rsi
pop rbp
%else
%if LIBVPX_YASM_WIN64
RESTORE_XMM
%endif
%endif
ret
%endmacro
;void vp8_copy32xn_sse3(
; unsigned char *src_ptr,
; int src_stride,
; unsigned char *dst_ptr,
; int dst_stride,
; int height);
global sym(vp8_copy32xn_sse3) PRIVATE
sym(vp8_copy32xn_sse3):
STACK_FRAME_CREATE_X3
.block_copy_sse3_loopx4:
lea end_ptr, [src_ptr+src_stride*2]
movdqu xmm0, XMMWORD PTR [src_ptr]
movdqu xmm1, XMMWORD PTR [src_ptr + 16]
movdqu xmm2, XMMWORD PTR [src_ptr + src_stride]
movdqu xmm3, XMMWORD PTR [src_ptr + src_stride + 16]
movdqu xmm4, XMMWORD PTR [end_ptr]
movdqu xmm5, XMMWORD PTR [end_ptr + 16]
movdqu xmm6, XMMWORD PTR [end_ptr + src_stride]
movdqu xmm7, XMMWORD PTR [end_ptr + src_stride + 16]
lea src_ptr, [src_ptr+src_stride*4]
lea end_ptr, [ref_ptr+ref_stride*2]
movdqa XMMWORD PTR [ref_ptr], xmm0
movdqa XMMWORD PTR [ref_ptr + 16], xmm1
movdqa XMMWORD PTR [ref_ptr + ref_stride], xmm2
movdqa XMMWORD PTR [ref_ptr + ref_stride + 16], xmm3
movdqa XMMWORD PTR [end_ptr], xmm4
movdqa XMMWORD PTR [end_ptr + 16], xmm5
movdqa XMMWORD PTR [end_ptr + ref_stride], xmm6
movdqa XMMWORD PTR [end_ptr + ref_stride + 16], xmm7
lea ref_ptr, [ref_ptr+ref_stride*4]
sub height, 4
cmp height, 4
jge .block_copy_sse3_loopx4
;Check to see if there is more rows need to be copied.
cmp height, 0
je .copy_is_done
.block_copy_sse3_loop:
movdqu xmm0, XMMWORD PTR [src_ptr]
movdqu xmm1, XMMWORD PTR [src_ptr + 16]
lea src_ptr, [src_ptr+src_stride]
movdqa XMMWORD PTR [ref_ptr], xmm0
movdqa XMMWORD PTR [ref_ptr + 16], xmm1
lea ref_ptr, [ref_ptr+ref_stride]
sub height, 1
jne .block_copy_sse3_loop
.copy_is_done:
STACK_FRAME_DESTROY_X3

16
media/libvpx/vp8/common/x86/idct_blk_mmx.c
Просмотреть файл

@ -36,7 +36,7 @@ void vp8_dequant_idct_add_y_block_mmx
else if (eobs[0] == 1)
{
vp8_dc_only_idct_add_mmx (q[0]*dq[0], dst, stride, dst, stride);
vpx_memset(q, 0, 2 * sizeof(q[0]));
memset(q, 0, 2 * sizeof(q[0]));
}
if (eobs[1] > 1)
@ -45,7 +45,7 @@ void vp8_dequant_idct_add_y_block_mmx
{
vp8_dc_only_idct_add_mmx (q[16]*dq[0], dst+4, stride,
dst+4, stride);
vpx_memset(q + 16, 0, 2 * sizeof(q[0]));
memset(q + 16, 0, 2 * sizeof(q[0]));
}
if (eobs[2] > 1)
@ -54,7 +54,7 @@ void vp8_dequant_idct_add_y_block_mmx
{
vp8_dc_only_idct_add_mmx (q[32]*dq[0], dst+8, stride,
dst+8, stride);
vpx_memset(q + 32, 0, 2 * sizeof(q[0]));
memset(q + 32, 0, 2 * sizeof(q[0]));
}
if (eobs[3] > 1)
@ -63,7 +63,7 @@ void vp8_dequant_idct_add_y_block_mmx
{
vp8_dc_only_idct_add_mmx (q[48]*dq[0], dst+12, stride,
dst+12, stride);
vpx_memset(q + 48, 0, 2 * sizeof(q[0]));
memset(q + 48, 0, 2 * sizeof(q[0]));
}
q += 64;
@ -85,7 +85,7 @@ void vp8_dequant_idct_add_uv_block_mmx
else if (eobs[0] == 1)
{
vp8_dc_only_idct_add_mmx (q[0]*dq[0], dstu, stride, dstu, stride);
vpx_memset(q, 0, 2 * sizeof(q[0]));
memset(q, 0, 2 * sizeof(q[0]));
}
if (eobs[1] > 1)
@ -94,7 +94,7 @@ void vp8_dequant_idct_add_uv_block_mmx
{
vp8_dc_only_idct_add_mmx (q[16]*dq[0], dstu+4, stride,
dstu+4, stride);
vpx_memset(q + 16, 0, 2 * sizeof(q[0]));
memset(q + 16, 0, 2 * sizeof(q[0]));
}
q += 32;
@ -109,7 +109,7 @@ void vp8_dequant_idct_add_uv_block_mmx
else if (eobs[0] == 1)
{
vp8_dc_only_idct_add_mmx (q[0]*dq[0], dstv, stride, dstv, stride);
vpx_memset(q, 0, 2 * sizeof(q[0]));
memset(q, 0, 2 * sizeof(q[0]));
}
if (eobs[1] > 1)
@ -118,7 +118,7 @@ void vp8_dequant_idct_add_uv_block_mmx
{
vp8_dc_only_idct_add_mmx (q[16]*dq[0], dstv+4, stride,
dstv+4, stride);
vpx_memset(q + 16, 0, 2 * sizeof(q[0]));
memset(q + 16, 0, 2 * sizeof(q[0]));
}
q += 32;

410
media/libvpx/vp8/common/x86/sad_sse2.asm
Просмотреть файл

@ -1,410 +0,0 @@
;
; Copyright (c) 2010 The WebM project authors. All Rights Reserved.
;
; Use of this source code is governed by a BSD-style license
; that can be found in the LICENSE file in the root of the source
; tree. An additional intellectual property rights grant can be found
; in the file PATENTS. All contributing project authors may
; be found in the AUTHORS file in the root of the source tree.
;
%include "vpx_ports/x86_abi_support.asm"
;unsigned int vp8_sad16x16_wmt(
; unsigned char *src_ptr,
; int src_stride,
; unsigned char *ref_ptr,
; int ref_stride)
global sym(vp8_sad16x16_wmt) PRIVATE
sym(vp8_sad16x16_wmt):
push rbp
mov rbp, rsp
SHADOW_ARGS_TO_STACK 4
SAVE_XMM 6
push rsi
push rdi
; end prolog
mov rsi, arg(0) ;src_ptr
mov rdi, arg(2) ;ref_ptr
movsxd rax, dword ptr arg(1) ;src_stride
movsxd rdx, dword ptr arg(3) ;ref_stride
lea rcx, [rsi+rax*8]
lea rcx, [rcx+rax*8]
pxor xmm6, xmm6
.x16x16sad_wmt_loop:
movq xmm0, QWORD PTR [rsi]
movq xmm2, QWORD PTR [rsi+8]
movq xmm1, QWORD PTR [rdi]
movq xmm3, QWORD PTR [rdi+8]
movq xmm4, QWORD PTR [rsi+rax]
movq xmm5, QWORD PTR [rdi+rdx]
punpcklbw xmm0, xmm2
punpcklbw xmm1, xmm3
psadbw xmm0, xmm1
movq xmm2, QWORD PTR [rsi+rax+8]
movq xmm3, QWORD PTR [rdi+rdx+8]
lea rsi, [rsi+rax*2]
lea rdi, [rdi+rdx*2]
punpcklbw xmm4, xmm2
punpcklbw xmm5, xmm3
psadbw xmm4, xmm5
paddw xmm6, xmm0
paddw xmm6, xmm4
cmp rsi, rcx
jne .x16x16sad_wmt_loop
movq xmm0, xmm6
psrldq xmm6, 8
paddw xmm0, xmm6
movq rax, xmm0
; begin epilog
pop rdi
pop rsi
RESTORE_XMM
UNSHADOW_ARGS
pop rbp
ret
;unsigned int vp8_sad8x16_wmt(
; unsigned char *src_ptr,
; int src_stride,
; unsigned char *ref_ptr,
; int ref_stride,
; int max_sad)
global sym(vp8_sad8x16_wmt) PRIVATE
sym(vp8_sad8x16_wmt):
push rbp
mov rbp, rsp
SHADOW_ARGS_TO_STACK 5
push rbx
push rsi
push rdi
; end prolog
mov rsi, arg(0) ;src_ptr
mov rdi, arg(2) ;ref_ptr
movsxd rbx, dword ptr arg(1) ;src_stride
movsxd rdx, dword ptr arg(3) ;ref_stride
lea rcx, [rsi+rbx*8]
lea rcx, [rcx+rbx*8]
pxor mm7, mm7
.x8x16sad_wmt_loop:
movq rax, mm7
cmp eax, arg(4)
ja .x8x16sad_wmt_early_exit
movq mm0, QWORD PTR [rsi]
movq mm1, QWORD PTR [rdi]
movq mm2, QWORD PTR [rsi+rbx]
movq mm3, QWORD PTR [rdi+rdx]
psadbw mm0, mm1
psadbw mm2, mm3
lea rsi, [rsi+rbx*2]
lea rdi, [rdi+rdx*2]
paddw mm7, mm0
paddw mm7, mm2
cmp rsi, rcx
jne .x8x16sad_wmt_loop
movq rax, mm7
.x8x16sad_wmt_early_exit:
; begin epilog
pop rdi
pop rsi
pop rbx
UNSHADOW_ARGS
pop rbp
ret
;unsigned int vp8_sad8x8_wmt(
; unsigned char *src_ptr,
; int src_stride,
; unsigned char *ref_ptr,
; int ref_stride)
global sym(vp8_sad8x8_wmt) PRIVATE
sym(vp8_sad8x8_wmt):
push rbp
mov rbp, rsp
SHADOW_ARGS_TO_STACK 5
push rbx
push rsi
push rdi
; end prolog
mov rsi, arg(0) ;src_ptr
mov rdi, arg(2) ;ref_ptr
movsxd rbx, dword ptr arg(1) ;src_stride
movsxd rdx, dword ptr arg(3) ;ref_stride
lea rcx, [rsi+rbx*8]
pxor mm7, mm7
.x8x8sad_wmt_loop:
movq rax, mm7
cmp eax, arg(4)
ja .x8x8sad_wmt_early_exit
movq mm0, QWORD PTR [rsi]
movq mm1, QWORD PTR [rdi]
psadbw mm0, mm1
lea rsi, [rsi+rbx]
add rdi, rdx
paddw mm7, mm0
cmp rsi, rcx
jne .x8x8sad_wmt_loop
movq rax, mm7
.x8x8sad_wmt_early_exit:
; begin epilog
pop rdi
pop rsi
pop rbx
UNSHADOW_ARGS
pop rbp
ret
;unsigned int vp8_sad4x4_wmt(
; unsigned char *src_ptr,
; int src_stride,
; unsigned char *ref_ptr,
; int ref_stride)
global sym(vp8_sad4x4_wmt) PRIVATE
sym(vp8_sad4x4_wmt):
push rbp
mov rbp, rsp
SHADOW_ARGS_TO_STACK 4
push rsi
push rdi
; end prolog
mov rsi, arg(0) ;src_ptr
mov rdi, arg(2) ;ref_ptr
movsxd rax, dword ptr arg(1) ;src_stride
movsxd rdx, dword ptr arg(3) ;ref_stride
movd mm0, DWORD PTR [rsi]
movd mm1, DWORD PTR [rdi]
movd mm2, DWORD PTR [rsi+rax]
movd mm3, DWORD PTR [rdi+rdx]
punpcklbw mm0, mm2
punpcklbw mm1, mm3
psadbw mm0, mm1
lea rsi, [rsi+rax*2]
lea rdi, [rdi+rdx*2]
movd mm4, DWORD PTR [rsi]
movd mm5, DWORD PTR [rdi]
movd mm6, DWORD PTR [rsi+rax]
movd mm7, DWORD PTR [rdi+rdx]
punpcklbw mm4, mm6
punpcklbw mm5, mm7
psadbw mm4, mm5
paddw mm0, mm4
movq rax, mm0
; begin epilog
pop rdi
pop rsi
UNSHADOW_ARGS
pop rbp
ret
;unsigned int vp8_sad16x8_wmt(
; unsigned char *src_ptr,
; int src_stride,
; unsigned char *ref_ptr,
; int ref_stride)
global sym(vp8_sad16x8_wmt) PRIVATE
sym(vp8_sad16x8_wmt):
push rbp
mov rbp, rsp
SHADOW_ARGS_TO_STACK 5
push rbx
push rsi
push rdi
; end prolog
mov rsi, arg(0) ;src_ptr
mov rdi, arg(2) ;ref_ptr
movsxd rbx, dword ptr arg(1) ;src_stride
movsxd rdx, dword ptr arg(3) ;ref_stride
lea rcx, [rsi+rbx*8]
pxor mm7, mm7
.x16x8sad_wmt_loop:
movq rax, mm7
cmp eax, arg(4)
ja .x16x8sad_wmt_early_exit
movq mm0, QWORD PTR [rsi]
movq mm2, QWORD PTR [rsi+8]
movq mm1, QWORD PTR [rdi]
movq mm3, QWORD PTR [rdi+8]
movq mm4, QWORD PTR [rsi+rbx]
movq mm5, QWORD PTR [rdi+rdx]
psadbw mm0, mm1
psadbw mm2, mm3
movq mm1, QWORD PTR [rsi+rbx+8]
movq mm3, QWORD PTR [rdi+rdx+8]
psadbw mm4, mm5
psadbw mm1, mm3
lea rsi, [rsi+rbx*2]
lea rdi, [rdi+rdx*2]
paddw mm0, mm2
paddw mm4, mm1
paddw mm7, mm0
paddw mm7, mm4
cmp rsi, rcx
jne .x16x8sad_wmt_loop
movq rax, mm7
.x16x8sad_wmt_early_exit:
; begin epilog
pop rdi
pop rsi
pop rbx
UNSHADOW_ARGS
pop rbp
ret
;void vp8_copy32xn_sse2(
; unsigned char *src_ptr,
; int src_stride,
; unsigned char *dst_ptr,
; int dst_stride,
; int height);
global sym(vp8_copy32xn_sse2) PRIVATE
sym(vp8_copy32xn_sse2):
push rbp
mov rbp, rsp
SHADOW_ARGS_TO_STACK 5
SAVE_XMM 7
push rsi
push rdi
; end prolog
mov rsi, arg(0) ;src_ptr
mov rdi, arg(2) ;dst_ptr
movsxd rax, dword ptr arg(1) ;src_stride
movsxd rdx, dword ptr arg(3) ;dst_stride
movsxd rcx, dword ptr arg(4) ;height
.block_copy_sse2_loopx4:
movdqu xmm0, XMMWORD PTR [rsi]
movdqu xmm1, XMMWORD PTR [rsi + 16]
movdqu xmm2, XMMWORD PTR [rsi + rax]
movdqu xmm3, XMMWORD PTR [rsi + rax + 16]
lea rsi, [rsi+rax*2]
movdqu xmm4, XMMWORD PTR [rsi]
movdqu xmm5, XMMWORD PTR [rsi + 16]
movdqu xmm6, XMMWORD PTR [rsi + rax]
movdqu xmm7, XMMWORD PTR [rsi + rax + 16]
lea rsi, [rsi+rax*2]
movdqa XMMWORD PTR [rdi], xmm0
movdqa XMMWORD PTR [rdi + 16], xmm1
movdqa XMMWORD PTR [rdi + rdx], xmm2
movdqa XMMWORD PTR [rdi + rdx + 16], xmm3
lea rdi, [rdi+rdx*2]
movdqa XMMWORD PTR [rdi], xmm4
movdqa XMMWORD PTR [rdi + 16], xmm5
movdqa XMMWORD PTR [rdi + rdx], xmm6
movdqa XMMWORD PTR [rdi + rdx + 16], xmm7
lea rdi, [rdi+rdx*2]
sub rcx, 4
cmp rcx, 4
jge .block_copy_sse2_loopx4
cmp rcx, 0
je .copy_is_done
.block_copy_sse2_loop:
movdqu xmm0, XMMWORD PTR [rsi]
movdqu xmm1, XMMWORD PTR [rsi + 16]
lea rsi, [rsi+rax]
movdqa XMMWORD PTR [rdi], xmm0
movdqa XMMWORD PTR [rdi + 16], xmm1
lea rdi, [rdi+rdx]
sub rcx, 1
jne .block_copy_sse2_loop
.copy_is_done:
; begin epilog
pop rdi
pop rsi
RESTORE_XMM
UNSHADOW_ARGS
pop rbp
ret

960
media/libvpx/vp8/common/x86/sad_sse3.asm
Просмотреть файл

@ -1,960 +0,0 @@
;
; Copyright (c) 2010 The WebM project authors. All Rights Reserved.
;
; Use of this source code is governed by a BSD-style license
; that can be found in the LICENSE file in the root of the source
; tree. An additional intellectual property rights grant can be found
; in the file PATENTS. All contributing project authors may
; be found in the AUTHORS file in the root of the source tree.
;
%include "vpx_ports/x86_abi_support.asm"
%macro STACK_FRAME_CREATE_X3 0
%if ABI_IS_32BIT
%define src_ptr rsi
%define src_stride rax
%define ref_ptr rdi
%define ref_stride rdx
%define end_ptr rcx
%define ret_var rbx
%define result_ptr arg(4)
%define max_sad arg(4)
%define height dword ptr arg(4)
push rbp
mov rbp, rsp
push rsi
push rdi
push rbx
mov rsi, arg(0) ; src_ptr
mov rdi, arg(2) ; ref_ptr
movsxd rax, dword ptr arg(1) ; src_stride
movsxd rdx, dword ptr arg(3) ; ref_stride
%else
%if LIBVPX_YASM_WIN64
SAVE_XMM 7, u
%define src_ptr rcx
%define src_stride rdx
%define ref_ptr r8
%define ref_stride r9
%define end_ptr r10
%define ret_var r11
%define result_ptr [rsp+xmm_stack_space+8+4*8]
%define max_sad [rsp+xmm_stack_space+8+4*8]
%define height dword ptr [rsp+xmm_stack_space+8+4*8]
%else
%define src_ptr rdi
%define src_stride rsi
%define ref_ptr rdx
%define ref_stride rcx
%define end_ptr r9
%define ret_var r10
%define result_ptr r8
%define max_sad r8
%define height r8
%endif
%endif
%endmacro
%macro STACK_FRAME_DESTROY_X3 0
%define src_ptr
%define src_stride
%define ref_ptr
%define ref_stride
%define end_ptr
%define ret_var
%define result_ptr
%define max_sad
%define height
%if ABI_IS_32BIT
pop rbx
pop rdi
pop rsi
pop rbp
%else
%if LIBVPX_YASM_WIN64
RESTORE_XMM
%endif
%endif
ret
%endmacro
%macro STACK_FRAME_CREATE_X4 0
%if ABI_IS_32BIT
%define src_ptr rsi
%define src_stride rax
%define r0_ptr rcx
%define r1_ptr rdx
%define r2_ptr rbx
%define r3_ptr rdi
%define ref_stride rbp
%define result_ptr arg(4)
push rbp
mov rbp, rsp
push rsi
push rdi
push rbx
push rbp
mov rdi, arg(2) ; ref_ptr_base
LOAD_X4_ADDRESSES rdi, rcx, rdx, rax, rdi
mov rsi, arg(0) ; src_ptr
movsxd rbx, dword ptr arg(1) ; src_stride
movsxd rbp, dword ptr arg(3) ; ref_stride
xchg rbx, rax
%else
%if LIBVPX_YASM_WIN64
SAVE_XMM 7, u
%define src_ptr rcx
%define src_stride rdx
%define r0_ptr rsi
%define r1_ptr r10
%define r2_ptr r11
%define r3_ptr r8
%define ref_stride r9
%define result_ptr [rsp+xmm_stack_space+16+4*8]
push rsi
LOAD_X4_ADDRESSES r8, r0_ptr, r1_ptr, r2_ptr, r3_ptr
%else
%define src_ptr rdi
%define src_stride rsi
%define r0_ptr r9
%define r1_ptr r10
%define r2_ptr r11
%define r3_ptr rdx
%define ref_stride rcx
%define result_ptr r8
LOAD_X4_ADDRESSES rdx, r0_ptr, r1_ptr, r2_ptr, r3_ptr
%endif
%endif
%endmacro
%macro STACK_FRAME_DESTROY_X4 0
%define src_ptr
%define src_stride
%define r0_ptr
%define r1_ptr
%define r2_ptr
%define r3_ptr
%define ref_stride
%define result_ptr
%if ABI_IS_32BIT
pop rbx
pop rdi
pop rsi
pop rbp
%else
%if LIBVPX_YASM_WIN64
pop rsi
RESTORE_XMM
%endif
%endif
ret
%endmacro
%macro PROCESS_16X2X3 5
%if %1==0
movdqa xmm0, XMMWORD PTR [%2]
lddqu xmm5, XMMWORD PTR [%3]
lddqu xmm6, XMMWORD PTR [%3+1]
lddqu xmm7, XMMWORD PTR [%3+2]
psadbw xmm5, xmm0
psadbw xmm6, xmm0
psadbw xmm7, xmm0
%else
movdqa xmm0, XMMWORD PTR [%2]
lddqu xmm1, XMMWORD PTR [%3]
lddqu xmm2, XMMWORD PTR [%3+1]
lddqu xmm3, XMMWORD PTR [%3+2]
psadbw xmm1, xmm0
psadbw xmm2, xmm0
psadbw xmm3, xmm0
paddw xmm5, xmm1
paddw xmm6, xmm2
paddw xmm7, xmm3
%endif
movdqa xmm0, XMMWORD PTR [%2+%4]
lddqu xmm1, XMMWORD PTR [%3+%5]
lddqu xmm2, XMMWORD PTR [%3+%5+1]
lddqu xmm3, XMMWORD PTR [%3+%5+2]
%if %1==0 || %1==1
lea %2, [%2+%4*2]
lea %3, [%3+%5*2]
%endif
psadbw xmm1, xmm0
psadbw xmm2, xmm0
psadbw xmm3, xmm0
paddw xmm5, xmm1
paddw xmm6, xmm2
paddw xmm7, xmm3
%endmacro
%macro PROCESS_8X2X3 5
%if %1==0
movq mm0, QWORD PTR [%2]
movq mm5, QWORD PTR [%3]
movq mm6, QWORD PTR [%3+1]
movq mm7, QWORD PTR [%3+2]
psadbw mm5, mm0
psadbw mm6, mm0
psadbw mm7, mm0
%else
movq mm0, QWORD PTR [%2]
movq mm1, QWORD PTR [%3]
movq mm2, QWORD PTR [%3+1]
movq mm3, QWORD PTR [%3+2]
psadbw mm1, mm0
psadbw mm2, mm0
psadbw mm3, mm0
paddw mm5, mm1
paddw mm6, mm2
paddw mm7, mm3
%endif
movq mm0, QWORD PTR [%2+%4]
movq mm1, QWORD PTR [%3+%5]
movq mm2, QWORD PTR [%3+%5+1]
movq mm3, QWORD PTR [%3+%5+2]
%if %1==0 || %1==1
lea %2, [%2+%4*2]
lea %3, [%3+%5*2]
%endif
psadbw mm1, mm0
psadbw mm2, mm0
psadbw mm3, mm0
paddw mm5, mm1
paddw mm6, mm2
paddw mm7, mm3
%endmacro
%macro LOAD_X4_ADDRESSES 5
mov %2, [%1+REG_SZ_BYTES*0]
mov %3, [%1+REG_SZ_BYTES*1]
mov %4, [%1+REG_SZ_BYTES*2]
mov %5, [%1+REG_SZ_BYTES*3]
%endmacro
%macro PROCESS_16X2X4 8
%if %1==0
movdqa xmm0, XMMWORD PTR [%2]
lddqu xmm4, XMMWORD PTR [%3]
lddqu xmm5, XMMWORD PTR [%4]
lddqu xmm6, XMMWORD PTR [%5]
lddqu xmm7, XMMWORD PTR [%6]
psadbw xmm4, xmm0
psadbw xmm5, xmm0
psadbw xmm6, xmm0
psadbw xmm7, xmm0
%else
movdqa xmm0, XMMWORD PTR [%2]
lddqu xmm1, XMMWORD PTR [%3]
lddqu xmm2, XMMWORD PTR [%4]
lddqu xmm3, XMMWORD PTR [%5]
psadbw xmm1, xmm0
psadbw xmm2, xmm0
psadbw xmm3, xmm0
paddw xmm4, xmm1
lddqu xmm1, XMMWORD PTR [%6]
paddw xmm5, xmm2
paddw xmm6, xmm3
psadbw xmm1, xmm0
paddw xmm7, xmm1
%endif
movdqa xmm0, XMMWORD PTR [%2+%7]
lddqu xmm1, XMMWORD PTR [%3+%8]
lddqu xmm2, XMMWORD PTR [%4+%8]
lddqu xmm3, XMMWORD PTR [%5+%8]
psadbw xmm1, xmm0
psadbw xmm2, xmm0
psadbw xmm3, xmm0
paddw xmm4, xmm1
lddqu xmm1, XMMWORD PTR [%6+%8]
paddw xmm5, xmm2
paddw xmm6, xmm3
%if %1==0 || %1==1
lea %2, [%2+%7*2]
lea %3, [%3+%8*2]
lea %4, [%4+%8*2]
lea %5, [%5+%8*2]
lea %6, [%6+%8*2]
%endif
psadbw xmm1, xmm0
paddw xmm7, xmm1
%endmacro
%macro PROCESS_8X2X4 8
%if %1==0
movq mm0, QWORD PTR [%2]
movq mm4, QWORD PTR [%3]
movq mm5, QWORD PTR [%4]
movq mm6, QWORD PTR [%5]
movq mm7, QWORD PTR [%6]
psadbw mm4, mm0
psadbw mm5, mm0
psadbw mm6, mm0
psadbw mm7, mm0
%else
movq mm0, QWORD PTR [%2]
movq mm1, QWORD PTR [%3]
movq mm2, QWORD PTR [%4]
movq mm3, QWORD PTR [%5]
psadbw mm1, mm0
psadbw mm2, mm0
psadbw mm3, mm0
paddw mm4, mm1
movq mm1, QWORD PTR [%6]
paddw mm5, mm2
paddw mm6, mm3
psadbw mm1, mm0
paddw mm7, mm1
%endif
movq mm0, QWORD PTR [%2+%7]
movq mm1, QWORD PTR [%3+%8]
movq mm2, QWORD PTR [%4+%8]
movq mm3, QWORD PTR [%5+%8]
psadbw mm1, mm0
psadbw mm2, mm0
psadbw mm3, mm0
paddw mm4, mm1
movq mm1, QWORD PTR [%6+%8]
paddw mm5, mm2
paddw mm6, mm3
%if %1==0 || %1==1
lea %2, [%2+%7*2]
lea %3, [%3+%8*2]
lea %4, [%4+%8*2]
lea %5, [%5+%8*2]
lea %6, [%6+%8*2]
%endif
psadbw mm1, mm0
paddw mm7, mm1
%endmacro
;void int vp8_sad16x16x3_sse3(
; unsigned char *src_ptr,
; int src_stride,
; unsigned char *ref_ptr,
; int ref_stride,
; int *results)
global sym(vp8_sad16x16x3_sse3) PRIVATE
sym(vp8_sad16x16x3_sse3):
STACK_FRAME_CREATE_X3
PROCESS_16X2X3 0, src_ptr, ref_ptr, src_stride, ref_stride
PROCESS_16X2X3 1, src_ptr, ref_ptr, src_stride, ref_stride
PROCESS_16X2X3 1, src_ptr, ref_ptr, src_stride, ref_stride
PROCESS_16X2X3 1, src_ptr, ref_ptr, src_stride, ref_stride
PROCESS_16X2X3 1, src_ptr, ref_ptr, src_stride, ref_stride
PROCESS_16X2X3 1, src_ptr, ref_ptr, src_stride, ref_stride
PROCESS_16X2X3 1, src_ptr, ref_ptr, src_stride, ref_stride
PROCESS_16X2X3 2, src_ptr, ref_ptr, src_stride, ref_stride
mov rcx, result_ptr
movq xmm0, xmm5
psrldq xmm5, 8
paddw xmm0, xmm5
movd [rcx], xmm0
;-
movq xmm0, xmm6
psrldq xmm6, 8
paddw xmm0, xmm6
movd [rcx+4], xmm0
;-
movq xmm0, xmm7
psrldq xmm7, 8
paddw xmm0, xmm7
movd [rcx+8], xmm0
STACK_FRAME_DESTROY_X3
;void int vp8_sad16x8x3_sse3(
; unsigned char *src_ptr,
; int src_stride,
; unsigned char *ref_ptr,
; int ref_stride,
; int *results)
global sym(vp8_sad16x8x3_sse3) PRIVATE
sym(vp8_sad16x8x3_sse3):
STACK_FRAME_CREATE_X3
PROCESS_16X2X3 0, src_ptr, ref_ptr, src_stride, ref_stride
PROCESS_16X2X3 1, src_ptr, ref_ptr, src_stride, ref_stride
PROCESS_16X2X3 1, src_ptr, ref_ptr, src_stride, ref_stride
PROCESS_16X2X3 2, src_ptr, ref_ptr, src_stride, ref_stride
mov rcx, result_ptr
movq xmm0, xmm5
psrldq xmm5, 8
paddw xmm0, xmm5
movd [rcx], xmm0
;-
movq xmm0, xmm6
psrldq xmm6, 8
paddw xmm0, xmm6
movd [rcx+4], xmm0
;-
movq xmm0, xmm7
psrldq xmm7, 8
paddw xmm0, xmm7
movd [rcx+8], xmm0
STACK_FRAME_DESTROY_X3
;void int vp8_sad8x16x3_sse3(
; unsigned char *src_ptr,
; int src_stride,
; unsigned char *ref_ptr,
; int ref_stride,
; int *results)
global sym(vp8_sad8x16x3_sse3) PRIVATE
sym(vp8_sad8x16x3_sse3):
STACK_FRAME_CREATE_X3
PROCESS_8X2X3 0, src_ptr, ref_ptr, src_stride, ref_stride
PROCESS_8X2X3 1, src_ptr, ref_ptr, src_stride, ref_stride
PROCESS_8X2X3 1, src_ptr, ref_ptr, src_stride, ref_stride
PROCESS_8X2X3 1, src_ptr, ref_ptr, src_stride, ref_stride
PROCESS_8X2X3 1, src_ptr, ref_ptr, src_stride, ref_stride
PROCESS_8X2X3 1, src_ptr, ref_ptr, src_stride, ref_stride
PROCESS_8X2X3 1, src_ptr, ref_ptr, src_stride, ref_stride
PROCESS_8X2X3 2, src_ptr, ref_ptr, src_stride, ref_stride
mov rcx, result_ptr
punpckldq mm5, mm6
movq [rcx], mm5
movd [rcx+8], mm7
STACK_FRAME_DESTROY_X3
;void int vp8_sad8x8x3_sse3(
; unsigned char *src_ptr,
; int src_stride,
; unsigned char *ref_ptr,
; int ref_stride,
; int *results)
global sym(vp8_sad8x8x3_sse3) PRIVATE
sym(vp8_sad8x8x3_sse3):
STACK_FRAME_CREATE_X3
PROCESS_8X2X3 0, src_ptr, ref_ptr, src_stride, ref_stride
PROCESS_8X2X3 1, src_ptr, ref_ptr, src_stride, ref_stride
PROCESS_8X2X3 1, src_ptr, ref_ptr, src_stride, ref_stride
PROCESS_8X2X3 2, src_ptr, ref_ptr, src_stride, ref_stride
mov rcx, result_ptr
punpckldq mm5, mm6
movq [rcx], mm5
movd [rcx+8], mm7
STACK_FRAME_DESTROY_X3
;void int vp8_sad4x4x3_sse3(
; unsigned char *src_ptr,
; int src_stride,
; unsigned char *ref_ptr,
; int ref_stride,
; int *results)
global sym(vp8_sad4x4x3_sse3) PRIVATE
sym(vp8_sad4x4x3_sse3):
STACK_FRAME_CREATE_X3
movd mm0, DWORD PTR [src_ptr]
movd mm1, DWORD PTR [ref_ptr]
movd mm2, DWORD PTR [src_ptr+src_stride]
movd mm3, DWORD PTR [ref_ptr+ref_stride]
punpcklbw mm0, mm2
punpcklbw mm1, mm3
movd mm4, DWORD PTR [ref_ptr+1]
movd mm5, DWORD PTR [ref_ptr+2]
movd mm2, DWORD PTR [ref_ptr+ref_stride+1]
movd mm3, DWORD PTR [ref_ptr+ref_stride+2]
psadbw mm1, mm0
punpcklbw mm4, mm2
punpcklbw mm5, mm3
psadbw mm4, mm0
psadbw mm5, mm0
lea src_ptr, [src_ptr+src_stride*2]
lea ref_ptr, [ref_ptr+ref_stride*2]
movd mm0, DWORD PTR [src_ptr]
movd mm2, DWORD PTR [ref_ptr]
movd mm3, DWORD PTR [src_ptr+src_stride]
movd mm6, DWORD PTR [ref_ptr+ref_stride]
punpcklbw mm0, mm3
punpcklbw mm2, mm6
movd mm3, DWORD PTR [ref_ptr+1]
movd mm7, DWORD PTR [ref_ptr+2]
psadbw mm2, mm0
paddw mm1, mm2
movd mm2, DWORD PTR [ref_ptr+ref_stride+1]
movd mm6, DWORD PTR [ref_ptr+ref_stride+2]
punpcklbw mm3, mm2
punpcklbw mm7, mm6
psadbw mm3, mm0
psadbw mm7, mm0
paddw mm3, mm4
paddw mm7, mm5
mov rcx, result_ptr
punpckldq mm1, mm3
movq [rcx], mm1
movd [rcx+8], mm7
STACK_FRAME_DESTROY_X3
;unsigned int vp8_sad16x16_sse3(
; unsigned char *src_ptr,
; int src_stride,
; unsigned char *ref_ptr,
; int ref_stride,
; int max_sad)
;%define lddqu movdqu
global sym(vp8_sad16x16_sse3) PRIVATE
sym(vp8_sad16x16_sse3):
STACK_FRAME_CREATE_X3
mov end_ptr, 4
pxor xmm7, xmm7
.vp8_sad16x16_sse3_loop:
movdqa xmm0, XMMWORD PTR [src_ptr]
movdqu xmm1, XMMWORD PTR [ref_ptr]
movdqa xmm2, XMMWORD PTR [src_ptr+src_stride]
movdqu xmm3, XMMWORD PTR [ref_ptr+ref_stride]
lea src_ptr, [src_ptr+src_stride*2]
lea ref_ptr, [ref_ptr+ref_stride*2]
movdqa xmm4, XMMWORD PTR [src_ptr]
movdqu xmm5, XMMWORD PTR [ref_ptr]
movdqa xmm6, XMMWORD PTR [src_ptr+src_stride]
psadbw xmm0, xmm1
movdqu xmm1, XMMWORD PTR [ref_ptr+ref_stride]
psadbw xmm2, xmm3
psadbw xmm4, xmm5
psadbw xmm6, xmm1
lea src_ptr, [src_ptr+src_stride*2]
lea ref_ptr, [ref_ptr+ref_stride*2]
paddw xmm7, xmm0
paddw xmm7, xmm2
paddw xmm7, xmm4
paddw xmm7, xmm6
sub end_ptr, 1
jne .vp8_sad16x16_sse3_loop
movq xmm0, xmm7
psrldq xmm7, 8
paddw xmm0, xmm7
movq rax, xmm0
STACK_FRAME_DESTROY_X3
;void vp8_copy32xn_sse3(
; unsigned char *src_ptr,
; int src_stride,
; unsigned char *dst_ptr,
; int dst_stride,
; int height);
global sym(vp8_copy32xn_sse3) PRIVATE
sym(vp8_copy32xn_sse3):
STACK_FRAME_CREATE_X3
.block_copy_sse3_loopx4:
lea end_ptr, [src_ptr+src_stride*2]
movdqu xmm0, XMMWORD PTR [src_ptr]
movdqu xmm1, XMMWORD PTR [src_ptr + 16]
movdqu xmm2, XMMWORD PTR [src_ptr + src_stride]
movdqu xmm3, XMMWORD PTR [src_ptr + src_stride + 16]
movdqu xmm4, XMMWORD PTR [end_ptr]
movdqu xmm5, XMMWORD PTR [end_ptr + 16]
movdqu xmm6, XMMWORD PTR [end_ptr + src_stride]
movdqu xmm7, XMMWORD PTR [end_ptr + src_stride + 16]
lea src_ptr, [src_ptr+src_stride*4]
lea end_ptr, [ref_ptr+ref_stride*2]
movdqa XMMWORD PTR [ref_ptr], xmm0
movdqa XMMWORD PTR [ref_ptr + 16], xmm1
movdqa XMMWORD PTR [ref_ptr + ref_stride], xmm2
movdqa XMMWORD PTR [ref_ptr + ref_stride + 16], xmm3
movdqa XMMWORD PTR [end_ptr], xmm4
movdqa XMMWORD PTR [end_ptr + 16], xmm5
movdqa XMMWORD PTR [end_ptr + ref_stride], xmm6
movdqa XMMWORD PTR [end_ptr + ref_stride + 16], xmm7
lea ref_ptr, [ref_ptr+ref_stride*4]
sub height, 4
cmp height, 4
jge .block_copy_sse3_loopx4
;Check to see if there is more rows need to be copied.
cmp height, 0
je .copy_is_done
.block_copy_sse3_loop:
movdqu xmm0, XMMWORD PTR [src_ptr]
movdqu xmm1, XMMWORD PTR [src_ptr + 16]
lea src_ptr, [src_ptr+src_stride]
movdqa XMMWORD PTR [ref_ptr], xmm0
movdqa XMMWORD PTR [ref_ptr + 16], xmm1
lea ref_ptr, [ref_ptr+ref_stride]
sub height, 1
jne .block_copy_sse3_loop
.copy_is_done:
STACK_FRAME_DESTROY_X3
;void vp8_sad16x16x4d_sse3(
; unsigned char *src_ptr,
; int src_stride,
; unsigned char *ref_ptr_base,
; int ref_stride,
; int *results)
global sym(vp8_sad16x16x4d_sse3) PRIVATE
sym(vp8_sad16x16x4d_sse3):
STACK_FRAME_CREATE_X4
PROCESS_16X2X4 0, src_ptr, r0_ptr, r1_ptr, r2_ptr, r3_ptr, src_stride, ref_stride
PROCESS_16X2X4 1, src_ptr, r0_ptr, r1_ptr, r2_ptr, r3_ptr, src_stride, ref_stride
PROCESS_16X2X4 1, src_ptr, r0_ptr, r1_ptr, r2_ptr, r3_ptr, src_stride, ref_stride
PROCESS_16X2X4 1, src_ptr, r0_ptr, r1_ptr, r2_ptr, r3_ptr, src_stride, ref_stride
PROCESS_16X2X4 1, src_ptr, r0_ptr, r1_ptr, r2_ptr, r3_ptr, src_stride, ref_stride
PROCESS_16X2X4 1, src_ptr, r0_ptr, r1_ptr, r2_ptr, r3_ptr, src_stride, ref_stride
PROCESS_16X2X4 1, src_ptr, r0_ptr, r1_ptr, r2_ptr, r3_ptr, src_stride, ref_stride
PROCESS_16X2X4 2, src_ptr, r0_ptr, r1_ptr, r2_ptr, r3_ptr, src_stride, ref_stride
%if ABI_IS_32BIT
pop rbp
%endif
mov rcx, result_ptr
movq xmm0, xmm4
psrldq xmm4, 8
paddw xmm0, xmm4
movd [rcx], xmm0
;-
movq xmm0, xmm5
psrldq xmm5, 8
paddw xmm0, xmm5
movd [rcx+4], xmm0
;-
movq xmm0, xmm6
psrldq xmm6, 8
paddw xmm0, xmm6
movd [rcx+8], xmm0
;-
movq xmm0, xmm7
psrldq xmm7, 8
paddw xmm0, xmm7
movd [rcx+12], xmm0
STACK_FRAME_DESTROY_X4
;void vp8_sad16x8x4d_sse3(
; unsigned char *src_ptr,
; int src_stride,
; unsigned char *ref_ptr_base,
; int ref_stride,
; int *results)
global sym(vp8_sad16x8x4d_sse3) PRIVATE
sym(vp8_sad16x8x4d_sse3):
STACK_FRAME_CREATE_X4
PROCESS_16X2X4 0, src_ptr, r0_ptr, r1_ptr, r2_ptr, r3_ptr, src_stride, ref_stride
PROCESS_16X2X4 1, src_ptr, r0_ptr, r1_ptr, r2_ptr, r3_ptr, src_stride, ref_stride
PROCESS_16X2X4 1, src_ptr, r0_ptr, r1_ptr, r2_ptr, r3_ptr, src_stride, ref_stride
PROCESS_16X2X4 2, src_ptr, r0_ptr, r1_ptr, r2_ptr, r3_ptr, src_stride, ref_stride
%if ABI_IS_32BIT
pop rbp
%endif
mov rcx, result_ptr
movq xmm0, xmm4
psrldq xmm4, 8
paddw xmm0, xmm4
movd [rcx], xmm0
;-
movq xmm0, xmm5
psrldq xmm5, 8
paddw xmm0, xmm5
movd [rcx+4], xmm0
;-
movq xmm0, xmm6
psrldq xmm6, 8
paddw xmm0, xmm6
movd [rcx+8], xmm0
;-
movq xmm0, xmm7
psrldq xmm7, 8
paddw xmm0, xmm7
movd [rcx+12], xmm0
STACK_FRAME_DESTROY_X4
;void int vp8_sad8x16x4d_sse3(
; unsigned char *src_ptr,
; int src_stride,
; unsigned char *ref_ptr,
; int ref_stride,
; int *results)
global sym(vp8_sad8x16x4d_sse3) PRIVATE
sym(vp8_sad8x16x4d_sse3):
STACK_FRAME_CREATE_X4
PROCESS_8X2X4 0, src_ptr, r0_ptr, r1_ptr, r2_ptr, r3_ptr, src_stride, ref_stride
PROCESS_8X2X4 1, src_ptr, r0_ptr, r1_ptr, r2_ptr, r3_ptr, src_stride, ref_stride
PROCESS_8X2X4 1, src_ptr, r0_ptr, r1_ptr, r2_ptr, r3_ptr, src_stride, ref_stride
PROCESS_8X2X4 1, src_ptr, r0_ptr, r1_ptr, r2_ptr, r3_ptr, src_stride, ref_stride
PROCESS_8X2X4 1, src_ptr, r0_ptr, r1_ptr, r2_ptr, r3_ptr, src_stride, ref_stride
PROCESS_8X2X4 1, src_ptr, r0_ptr, r1_ptr, r2_ptr, r3_ptr, src_stride, ref_stride
PROCESS_8X2X4 1, src_ptr, r0_ptr, r1_ptr, r2_ptr, r3_ptr, src_stride, ref_stride
PROCESS_8X2X4 2, src_ptr, r0_ptr, r1_ptr, r2_ptr, r3_ptr, src_stride, ref_stride
%if ABI_IS_32BIT
pop rbp
%endif
mov rcx, result_ptr
punpckldq mm4, mm5
punpckldq mm6, mm7
movq [rcx], mm4
movq [rcx+8], mm6
STACK_FRAME_DESTROY_X4
;void int vp8_sad8x8x4d_sse3(
; unsigned char *src_ptr,
; int src_stride,
; unsigned char *ref_ptr,
; int ref_stride,
; int *results)
global sym(vp8_sad8x8x4d_sse3) PRIVATE
sym(vp8_sad8x8x4d_sse3):
STACK_FRAME_CREATE_X4
PROCESS_8X2X4 0, src_ptr, r0_ptr, r1_ptr, r2_ptr, r3_ptr, src_stride, ref_stride
PROCESS_8X2X4 1, src_ptr, r0_ptr, r1_ptr, r2_ptr, r3_ptr, src_stride, ref_stride
PROCESS_8X2X4 1, src_ptr, r0_ptr, r1_ptr, r2_ptr, r3_ptr, src_stride, ref_stride
PROCESS_8X2X4 2, src_ptr, r0_ptr, r1_ptr, r2_ptr, r3_ptr, src_stride, ref_stride
%if ABI_IS_32BIT
pop rbp
%endif
mov rcx, result_ptr
punpckldq mm4, mm5
punpckldq mm6, mm7
movq [rcx], mm4
movq [rcx+8], mm6
STACK_FRAME_DESTROY_X4
;void int vp8_sad4x4x4d_sse3(
; unsigned char *src_ptr,
; int src_stride,
; unsigned char *ref_ptr,
; int ref_stride,
; int *results)
global sym(vp8_sad4x4x4d_sse3) PRIVATE
sym(vp8_sad4x4x4d_sse3):
STACK_FRAME_CREATE_X4
movd mm0, DWORD PTR [src_ptr]
movd mm1, DWORD PTR [r0_ptr]
movd mm2, DWORD PTR [src_ptr+src_stride]
movd mm3, DWORD PTR [r0_ptr+ref_stride]
punpcklbw mm0, mm2
punpcklbw mm1, mm3
movd mm4, DWORD PTR [r1_ptr]
movd mm5, DWORD PTR [r2_ptr]
movd mm6, DWORD PTR [r3_ptr]
movd mm2, DWORD PTR [r1_ptr+ref_stride]
movd mm3, DWORD PTR [r2_ptr+ref_stride]
movd mm7, DWORD PTR [r3_ptr+ref_stride]
psadbw mm1, mm0
punpcklbw mm4, mm2
punpcklbw mm5, mm3
punpcklbw mm6, mm7
psadbw mm4, mm0
psadbw mm5, mm0
psadbw mm6, mm0
lea src_ptr, [src_ptr+src_stride*2]
lea r0_ptr, [r0_ptr+ref_stride*2]
lea r1_ptr, [r1_ptr+ref_stride*2]
lea r2_ptr, [r2_ptr+ref_stride*2]
lea r3_ptr, [r3_ptr+ref_stride*2]
movd mm0, DWORD PTR [src_ptr]
movd mm2, DWORD PTR [r0_ptr]
movd mm3, DWORD PTR [src_ptr+src_stride]
movd mm7, DWORD PTR [r0_ptr+ref_stride]
punpcklbw mm0, mm3
punpcklbw mm2, mm7
movd mm3, DWORD PTR [r1_ptr]
movd mm7, DWORD PTR [r2_ptr]
psadbw mm2, mm0
%if ABI_IS_32BIT
mov rax, rbp
pop rbp
%define ref_stride rax
%endif
mov rsi, result_ptr
paddw mm1, mm2
movd [rsi], mm1
movd mm2, DWORD PTR [r1_ptr+ref_stride]
movd mm1, DWORD PTR [r2_ptr+ref_stride]
punpcklbw mm3, mm2
punpcklbw mm7, mm1
psadbw mm3, mm0
psadbw mm7, mm0
movd mm2, DWORD PTR [r3_ptr]
movd mm1, DWORD PTR [r3_ptr+ref_stride]
paddw mm3, mm4
paddw mm7, mm5
movd [rsi+4], mm3
punpcklbw mm2, mm1
movd [rsi+8], mm7
psadbw mm2, mm0
paddw mm2, mm6
movd [rsi+12], mm2
STACK_FRAME_DESTROY_X4

353
media/libvpx/vp8/common/x86/sad_sse4.asm
Просмотреть файл

@ -1,353 +0,0 @@
;
; Copyright (c) 2010 The WebM project authors. All Rights Reserved.
;
; Use of this source code is governed by a BSD-style license
; that can be found in the LICENSE file in the root of the source
; tree. An additional intellectual property rights grant can be found
; in the file PATENTS. All contributing project authors may
; be found in the AUTHORS file in the root of the source tree.
;
%include "vpx_ports/x86_abi_support.asm"
%macro PROCESS_16X2X8 1
%if %1
movdqa xmm0, XMMWORD PTR [rsi]
movq xmm1, MMWORD PTR [rdi]
movq xmm3, MMWORD PTR [rdi+8]
movq xmm2, MMWORD PTR [rdi+16]
punpcklqdq xmm1, xmm3
punpcklqdq xmm3, xmm2
movdqa xmm2, xmm1
mpsadbw xmm1, xmm0, 0x0
mpsadbw xmm2, xmm0, 0x5
psrldq xmm0, 8
movdqa xmm4, xmm3
mpsadbw xmm3, xmm0, 0x0
mpsadbw xmm4, xmm0, 0x5
paddw xmm1, xmm2
paddw xmm1, xmm3
paddw xmm1, xmm4
%else
movdqa xmm0, XMMWORD PTR [rsi]
movq xmm5, MMWORD PTR [rdi]
movq xmm3, MMWORD PTR [rdi+8]
movq xmm2, MMWORD PTR [rdi+16]
punpcklqdq xmm5, xmm3
punpcklqdq xmm3, xmm2
movdqa xmm2, xmm5
mpsadbw xmm5, xmm0, 0x0
mpsadbw xmm2, xmm0, 0x5
psrldq xmm0, 8
movdqa xmm4, xmm3
mpsadbw xmm3, xmm0, 0x0
mpsadbw xmm4, xmm0, 0x5
paddw xmm5, xmm2
paddw xmm5, xmm3
paddw xmm5, xmm4
paddw xmm1, xmm5
%endif
movdqa xmm0, XMMWORD PTR [rsi + rax]
movq xmm5, MMWORD PTR [rdi+ rdx]
movq xmm3, MMWORD PTR [rdi+ rdx+8]
movq xmm2, MMWORD PTR [rdi+ rdx+16]
punpcklqdq xmm5, xmm3
punpcklqdq xmm3, xmm2
lea rsi, [rsi+rax*2]
lea rdi, [rdi+rdx*2]
movdqa xmm2, xmm5
mpsadbw xmm5, xmm0, 0x0
mpsadbw xmm2, xmm0, 0x5
psrldq xmm0, 8
movdqa xmm4, xmm3
mpsadbw xmm3, xmm0, 0x0
mpsadbw xmm4, xmm0, 0x5
paddw xmm5, xmm2
paddw xmm5, xmm3
paddw xmm5, xmm4
paddw xmm1, xmm5
%endmacro
%macro PROCESS_8X2X8 1
%if %1
movq xmm0, MMWORD PTR [rsi]
movq xmm1, MMWORD PTR [rdi]
movq xmm3, MMWORD PTR [rdi+8]
punpcklqdq xmm1, xmm3
movdqa xmm2, xmm1
mpsadbw xmm1, xmm0, 0x0
mpsadbw xmm2, xmm0, 0x5
paddw xmm1, xmm2
%else
movq xmm0, MMWORD PTR [rsi]
movq xmm5, MMWORD PTR [rdi]
movq xmm3, MMWORD PTR [rdi+8]
punpcklqdq xmm5, xmm3
movdqa xmm2, xmm5
mpsadbw xmm5, xmm0, 0x0
mpsadbw xmm2, xmm0, 0x5
paddw xmm5, xmm2
paddw xmm1, xmm5
%endif
movq xmm0, MMWORD PTR [rsi + rax]
movq xmm5, MMWORD PTR [rdi+ rdx]
movq xmm3, MMWORD PTR [rdi+ rdx+8]
punpcklqdq xmm5, xmm3
lea rsi, [rsi+rax*2]
lea rdi, [rdi+rdx*2]
movdqa xmm2, xmm5
mpsadbw xmm5, xmm0, 0x0
mpsadbw xmm2, xmm0, 0x5
paddw xmm5, xmm2
paddw xmm1, xmm5
%endmacro
%macro PROCESS_4X2X8 1
%if %1
movd xmm0, [rsi]
movq xmm1, MMWORD PTR [rdi]
movq xmm3, MMWORD PTR [rdi+8]
punpcklqdq xmm1, xmm3
mpsadbw xmm1, xmm0, 0x0
%else
movd xmm0, [rsi]
movq xmm5, MMWORD PTR [rdi]
movq xmm3, MMWORD PTR [rdi+8]
punpcklqdq xmm5, xmm3
mpsadbw xmm5, xmm0, 0x0
paddw xmm1, xmm5
%endif
movd xmm0, [rsi + rax]
movq xmm5, MMWORD PTR [rdi+ rdx]
movq xmm3, MMWORD PTR [rdi+ rdx+8]
punpcklqdq xmm5, xmm3
lea rsi, [rsi+rax*2]
lea rdi, [rdi+rdx*2]
mpsadbw xmm5, xmm0, 0x0
paddw xmm1, xmm5
%endmacro
;void vp8_sad16x16x8_sse4(
; const unsigned char *src_ptr,
; int src_stride,
; const unsigned char *ref_ptr,
; int ref_stride,
; unsigned short *sad_array);
global sym(vp8_sad16x16x8_sse4) PRIVATE
sym(vp8_sad16x16x8_sse4):
push rbp
mov rbp, rsp
SHADOW_ARGS_TO_STACK 5
push rsi
push rdi
; end prolog
mov rsi, arg(0) ;src_ptr
mov rdi, arg(2) ;ref_ptr
movsxd rax, dword ptr arg(1) ;src_stride
movsxd rdx, dword ptr arg(3) ;ref_stride
PROCESS_16X2X8 1
PROCESS_16X2X8 0
PROCESS_16X2X8 0
PROCESS_16X2X8 0
PROCESS_16X2X8 0
PROCESS_16X2X8 0
PROCESS_16X2X8 0
PROCESS_16X2X8 0
mov rdi, arg(4) ;Results
movdqa XMMWORD PTR [rdi], xmm1
; begin epilog
pop rdi
pop rsi
UNSHADOW_ARGS
pop rbp
ret
;void vp8_sad16x8x8_sse4(
; const unsigned char *src_ptr,
; int src_stride,
; const unsigned char *ref_ptr,
; int ref_stride,
; unsigned short *sad_array
;);
global sym(vp8_sad16x8x8_sse4) PRIVATE
sym(vp8_sad16x8x8_sse4):
push rbp
mov rbp, rsp
SHADOW_ARGS_TO_STACK 5
push rsi
push rdi
; end prolog
mov rsi, arg(0) ;src_ptr
mov rdi, arg(2) ;ref_ptr
movsxd rax, dword ptr arg(1) ;src_stride
movsxd rdx, dword ptr arg(3) ;ref_stride
PROCESS_16X2X8 1
PROCESS_16X2X8 0
PROCESS_16X2X8 0
PROCESS_16X2X8 0
mov rdi, arg(4) ;Results
movdqa XMMWORD PTR [rdi], xmm1
; begin epilog
pop rdi
pop rsi
UNSHADOW_ARGS
pop rbp
ret
;void vp8_sad8x8x8_sse4(
; const unsigned char *src_ptr,
; int src_stride,
; const unsigned char *ref_ptr,
; int ref_stride,
; unsigned short *sad_array
;);
global sym(vp8_sad8x8x8_sse4) PRIVATE
sym(vp8_sad8x8x8_sse4):
push rbp
mov rbp, rsp
SHADOW_ARGS_TO_STACK 5
push rsi
push rdi
; end prolog
mov rsi, arg(0) ;src_ptr
mov rdi, arg(2) ;ref_ptr
movsxd rax, dword ptr arg(1) ;src_stride
movsxd rdx, dword ptr arg(3) ;ref_stride
PROCESS_8X2X8 1
PROCESS_8X2X8 0
PROCESS_8X2X8 0
PROCESS_8X2X8 0
mov rdi, arg(4) ;Results
movdqa XMMWORD PTR [rdi], xmm1
; begin epilog
pop rdi
pop rsi
UNSHADOW_ARGS
pop rbp
ret
;void vp8_sad8x16x8_sse4(
; const unsigned char *src_ptr,
; int src_stride,
; const unsigned char *ref_ptr,
; int ref_stride,
; unsigned short *sad_array
;);
global sym(vp8_sad8x16x8_sse4) PRIVATE
sym(vp8_sad8x16x8_sse4):
push rbp
mov rbp, rsp
SHADOW_ARGS_TO_STACK 5
push rsi
push rdi
; end prolog
mov rsi, arg(0) ;src_ptr
mov rdi, arg(2) ;ref_ptr
movsxd rax, dword ptr arg(1) ;src_stride
movsxd rdx, dword ptr arg(3) ;ref_stride
PROCESS_8X2X8 1
PROCESS_8X2X8 0
PROCESS_8X2X8 0
PROCESS_8X2X8 0
PROCESS_8X2X8 0
PROCESS_8X2X8 0
PROCESS_8X2X8 0
PROCESS_8X2X8 0
mov rdi, arg(4) ;Results
movdqa XMMWORD PTR [rdi], xmm1
; begin epilog
pop rdi
pop rsi
UNSHADOW_ARGS
pop rbp
ret
;void vp8_sad4x4x8_c(
; const unsigned char *src_ptr,
; int src_stride,
; const unsigned char *ref_ptr,
; int ref_stride,
; unsigned short *sad_array
;);
global sym(vp8_sad4x4x8_sse4) PRIVATE
sym(vp8_sad4x4x8_sse4):
push rbp
mov rbp, rsp
SHADOW_ARGS_TO_STACK 5
push rsi
push rdi
; end prolog
mov rsi, arg(0) ;src_ptr
mov rdi, arg(2) ;ref_ptr
movsxd rax, dword ptr arg(1) ;src_stride
movsxd rdx, dword ptr arg(3) ;ref_stride
PROCESS_4X2X8 1
PROCESS_4X2X8 0
mov rdi, arg(4) ;Results
movdqa XMMWORD PTR [rdi], xmm1
; begin epilog
pop rdi
pop rsi
UNSHADOW_ARGS
pop rbp
ret

387
media/libvpx/vp8/common/x86/variance_impl_sse2.asm
Просмотреть файл

@ -13,393 +13,6 @@
%define xmm_filter_shift 7
;unsigned int vp8_get_mb_ss_sse2
;(
; short *src_ptr
;)
global sym(vp8_get_mb_ss_sse2) PRIVATE
sym(vp8_get_mb_ss_sse2):
push rbp
mov rbp, rsp
SHADOW_ARGS_TO_STACK 1
GET_GOT rbx
push rsi
push rdi
sub rsp, 16
; end prolog
mov rax, arg(0) ;[src_ptr]
mov rcx, 8
pxor xmm4, xmm4
.NEXTROW:
movdqa xmm0, [rax]
movdqa xmm1, [rax+16]
movdqa xmm2, [rax+32]
movdqa xmm3, [rax+48]
pmaddwd xmm0, xmm0
pmaddwd xmm1, xmm1
pmaddwd xmm2, xmm2
pmaddwd xmm3, xmm3
paddd xmm0, xmm1
paddd xmm2, xmm3
paddd xmm4, xmm0
paddd xmm4, xmm2
add rax, 0x40
dec rcx
ja .NEXTROW
movdqa xmm3,xmm4
psrldq xmm4,8
paddd xmm4,xmm3
movdqa xmm3,xmm4
psrldq xmm4,4
paddd xmm4,xmm3
movq rax,xmm4
; begin epilog
add rsp, 16
pop rdi
pop rsi
RESTORE_GOT
UNSHADOW_ARGS
pop rbp
ret
;unsigned int vp8_get16x16var_sse2
;(
; unsigned char * src_ptr,
; int source_stride,
; unsigned char * ref_ptr,
; int recon_stride,
; unsigned int * SSE,
; int * Sum
;)
global sym(vp8_get16x16var_sse2) PRIVATE
sym(vp8_get16x16var_sse2):
push rbp
mov rbp, rsp
SHADOW_ARGS_TO_STACK 6
SAVE_XMM 7
push rbx
push rsi
push rdi
; end prolog
mov rsi, arg(0) ;[src_ptr]
mov rdi, arg(2) ;[ref_ptr]
movsxd rax, DWORD PTR arg(1) ;[source_stride]
movsxd rdx, DWORD PTR arg(3) ;[recon_stride]
; Prefetch data
lea rcx, [rax+rax*2]
prefetcht0 [rsi]
prefetcht0 [rsi+rax]
prefetcht0 [rsi+rax*2]
prefetcht0 [rsi+rcx]
lea rbx, [rsi+rax*4]
prefetcht0 [rbx]
prefetcht0 [rbx+rax]
prefetcht0 [rbx+rax*2]
prefetcht0 [rbx+rcx]
lea rcx, [rdx+rdx*2]
prefetcht0 [rdi]
prefetcht0 [rdi+rdx]
prefetcht0 [rdi+rdx*2]
prefetcht0 [rdi+rcx]
lea rbx, [rdi+rdx*4]
prefetcht0 [rbx]
prefetcht0 [rbx+rdx]
prefetcht0 [rbx+rdx*2]
prefetcht0 [rbx+rcx]
pxor xmm0, xmm0 ; clear xmm0 for unpack
pxor xmm7, xmm7 ; clear xmm7 for accumulating diffs
pxor xmm6, xmm6 ; clear xmm6 for accumulating sse
mov rcx, 16
.var16loop:
movdqu xmm1, XMMWORD PTR [rsi]
movdqu xmm2, XMMWORD PTR [rdi]
prefetcht0 [rsi+rax*8]
prefetcht0 [rdi+rdx*8]
movdqa xmm3, xmm1
movdqa xmm4, xmm2
punpcklbw xmm1, xmm0
punpckhbw xmm3, xmm0
punpcklbw xmm2, xmm0
punpckhbw xmm4, xmm0
psubw xmm1, xmm2
psubw xmm3, xmm4
paddw xmm7, xmm1
pmaddwd xmm1, xmm1
paddw xmm7, xmm3
pmaddwd xmm3, xmm3
paddd xmm6, xmm1
paddd xmm6, xmm3
add rsi, rax
add rdi, rdx
sub rcx, 1
jnz .var16loop
movdqa xmm1, xmm6
pxor xmm6, xmm6
pxor xmm5, xmm5
punpcklwd xmm6, xmm7
punpckhwd xmm5, xmm7
psrad xmm5, 16
psrad xmm6, 16
paddd xmm6, xmm5
movdqa xmm2, xmm1
punpckldq xmm1, xmm0
punpckhdq xmm2, xmm0
movdqa xmm7, xmm6
paddd xmm1, xmm2
punpckldq xmm6, xmm0
punpckhdq xmm7, xmm0
paddd xmm6, xmm7
movdqa xmm2, xmm1
movdqa xmm7, xmm6
psrldq xmm1, 8
psrldq xmm6, 8
paddd xmm7, xmm6
paddd xmm1, xmm2
mov rax, arg(5) ;[Sum]
mov rdi, arg(4) ;[SSE]
movd DWORD PTR [rax], xmm7
movd DWORD PTR [rdi], xmm1
; begin epilog
pop rdi
pop rsi
pop rbx
RESTORE_XMM
UNSHADOW_ARGS
pop rbp
ret
;unsigned int vp8_get8x8var_sse2
;(
; unsigned char * src_ptr,
; int source_stride,
; unsigned char * ref_ptr,
; int recon_stride,
; unsigned int * SSE,
; int * Sum
;)
global sym(vp8_get8x8var_sse2) PRIVATE
sym(vp8_get8x8var_sse2):
push rbp
mov rbp, rsp
SHADOW_ARGS_TO_STACK 6
SAVE_XMM 7
GET_GOT rbx
push rsi
push rdi
sub rsp, 16
; end prolog
mov rsi, arg(0) ;[src_ptr]
mov rdi, arg(2) ;[ref_ptr]
movsxd rax, DWORD PTR arg(1) ;[source_stride]
movsxd rdx, DWORD PTR arg(3) ;[recon_stride]
pxor xmm0, xmm0 ; clear xmm0 for unpack
pxor xmm7, xmm7 ; clear xmm7 for accumulating diffs
movq xmm1, QWORD PTR [rsi]
movq xmm2, QWORD PTR [rdi]
punpcklbw xmm1, xmm0
punpcklbw xmm2, xmm0
psubsw xmm1, xmm2
paddw xmm7, xmm1
pmaddwd xmm1, xmm1
movq xmm2, QWORD PTR[rsi + rax]
movq xmm3, QWORD PTR[rdi + rdx]
punpcklbw xmm2, xmm0
punpcklbw xmm3, xmm0
psubsw xmm2, xmm3
paddw xmm7, xmm2
pmaddwd xmm2, xmm2
paddd xmm1, xmm2
movq xmm2, QWORD PTR[rsi + rax * 2]
movq xmm3, QWORD PTR[rdi + rdx * 2]
punpcklbw xmm2, xmm0
punpcklbw xmm3, xmm0
psubsw xmm2, xmm3
paddw xmm7, xmm2
pmaddwd xmm2, xmm2
paddd xmm1, xmm2
lea rsi, [rsi + rax * 2]
lea rdi, [rdi + rdx * 2]
movq xmm2, QWORD PTR[rsi + rax]
movq xmm3, QWORD PTR[rdi + rdx]
punpcklbw xmm2, xmm0
punpcklbw xmm3, xmm0
psubsw xmm2, xmm3
paddw xmm7, xmm2
pmaddwd xmm2, xmm2
paddd xmm1, xmm2
movq xmm2, QWORD PTR[rsi + rax *2]
movq xmm3, QWORD PTR[rdi + rdx *2]
punpcklbw xmm2, xmm0
punpcklbw xmm3, xmm0
psubsw xmm2, xmm3
paddw xmm7, xmm2
pmaddwd xmm2, xmm2
paddd xmm1, xmm2
lea rsi, [rsi + rax * 2]
lea rdi, [rdi + rdx * 2]
movq xmm2, QWORD PTR[rsi + rax]
movq xmm3, QWORD PTR[rdi + rdx]
punpcklbw xmm2, xmm0
punpcklbw xmm3, xmm0
psubsw xmm2, xmm3
paddw xmm7, xmm2
pmaddwd xmm2, xmm2
paddd xmm1, xmm2
movq xmm2, QWORD PTR[rsi + rax *2]
movq xmm3, QWORD PTR[rdi + rdx *2]
punpcklbw xmm2, xmm0
punpcklbw xmm3, xmm0
psubsw xmm2, xmm3
paddw xmm7, xmm2
pmaddwd xmm2, xmm2
paddd xmm1, xmm2
lea rsi, [rsi + rax * 2]
lea rdi, [rdi + rdx * 2]
movq xmm2, QWORD PTR[rsi + rax]
movq xmm3, QWORD PTR[rdi + rdx]
punpcklbw xmm2, xmm0
punpcklbw xmm3, xmm0
psubsw xmm2, xmm3
paddw xmm7, xmm2
pmaddwd xmm2, xmm2
paddd xmm1, xmm2
movdqa xmm6, xmm7
punpcklwd xmm6, xmm0
punpckhwd xmm7, xmm0
movdqa xmm2, xmm1
paddw xmm6, xmm7
punpckldq xmm1, xmm0
punpckhdq xmm2, xmm0
movdqa xmm7, xmm6
paddd xmm1, xmm2
punpckldq xmm6, xmm0
punpckhdq xmm7, xmm0
paddw xmm6, xmm7
movdqa xmm2, xmm1
movdqa xmm7, xmm6
psrldq xmm1, 8
psrldq xmm6, 8
paddw xmm7, xmm6
paddd xmm1, xmm2
mov rax, arg(5) ;[Sum]
mov rdi, arg(4) ;[SSE]
movq rdx, xmm7
movsx rcx, dx
mov dword ptr [rax], ecx
movd DWORD PTR [rdi], xmm1
; begin epilog
add rsp, 16
pop rdi
pop rsi
RESTORE_GOT
RESTORE_XMM
UNSHADOW_ARGS
pop rbp
ret
;void vp8_filter_block2d_bil_var_sse2
;(
; unsigned char *ref_ptr,

10
media/libvpx/vp8/common/x86/variance_ssse3.c
Просмотреть файл

@ -8,19 +8,11 @@
* be found in the AUTHORS file in the root of the source tree.
*/
#include "./vp8_rtcd.h"
#include "vpx_config.h"
#include "vp8/common/variance.h"
#include "vpx_ports/mem.h"
extern unsigned int vp8_get16x16var_sse2
(
const unsigned char *src_ptr,
int source_stride,
const unsigned char *ref_ptr,
int recon_stride,
unsigned int *SSE,
int *Sum
);
extern void vp8_half_horiz_vert_variance16x_h_sse2
(
const unsigned char *ref_ptr,

22
media/libvpx/vp8/common/x86/vp8_asm_stubs.c
Просмотреть файл

@ -127,7 +127,7 @@ void vp8_sixtap_predict4x4_mmx
int dst_pitch
)
{
DECLARE_ALIGNED_ARRAY(16, unsigned short, FData2, 16*16); /* Temp data bufffer used in filtering */
DECLARE_ALIGNED(16, unsigned short, FData2[16*16]); /* Temp data bufffer used in filtering */
const short *HFilter, *VFilter;
HFilter = vp8_six_tap_mmx[xoffset];
vp8_filter_block1d_h6_mmx(src_ptr - (2 * src_pixels_per_line), FData2, src_pixels_per_line, 1, 9, 8, HFilter);
@ -148,7 +148,7 @@ void vp8_sixtap_predict16x16_mmx
)
{
DECLARE_ALIGNED_ARRAY(16, unsigned short, FData2, 24*24); /* Temp data bufffer used in filtering */
DECLARE_ALIGNED(16, unsigned short, FData2[24*24]); /* Temp data bufffer used in filtering */
const short *HFilter, *VFilter;
@ -180,7 +180,7 @@ void vp8_sixtap_predict8x8_mmx
)
{
DECLARE_ALIGNED_ARRAY(16, unsigned short, FData2, 256); /* Temp data bufffer used in filtering */
DECLARE_ALIGNED(16, unsigned short, FData2[256]); /* Temp data bufffer used in filtering */
const short *HFilter, *VFilter;
@ -206,7 +206,7 @@ void vp8_sixtap_predict8x4_mmx
)
{
DECLARE_ALIGNED_ARRAY(16, unsigned short, FData2, 256); /* Temp data bufffer used in filtering */
DECLARE_ALIGNED(16, unsigned short, FData2[256]); /* Temp data bufffer used in filtering */
const short *HFilter, *VFilter;
@ -252,7 +252,7 @@ void vp8_sixtap_predict16x16_sse2
)
{
DECLARE_ALIGNED_ARRAY(16, unsigned short, FData2, 24*24); /* Temp data bufffer used in filtering */
DECLARE_ALIGNED(16, unsigned short, FData2[24*24]); /* Temp data bufffer used in filtering */
const short *HFilter, *VFilter;
@ -292,7 +292,7 @@ void vp8_sixtap_predict8x8_sse2
int dst_pitch
)
{
DECLARE_ALIGNED_ARRAY(16, unsigned short, FData2, 256); /* Temp data bufffer used in filtering */
DECLARE_ALIGNED(16, unsigned short, FData2[256]); /* Temp data bufffer used in filtering */
const short *HFilter, *VFilter;
if (xoffset)
@ -330,7 +330,7 @@ void vp8_sixtap_predict8x4_sse2
int dst_pitch
)
{
DECLARE_ALIGNED_ARRAY(16, unsigned short, FData2, 256); /* Temp data bufffer used in filtering */
DECLARE_ALIGNED(16, unsigned short, FData2[256]); /* Temp data bufffer used in filtering */
const short *HFilter, *VFilter;
if (xoffset)
@ -432,7 +432,7 @@ void vp8_sixtap_predict16x16_ssse3
)
{
DECLARE_ALIGNED_ARRAY(16, unsigned char, FData2, 24*24);
DECLARE_ALIGNED(16, unsigned char, FData2[24*24]);
if (xoffset)
{
@ -480,7 +480,7 @@ void vp8_sixtap_predict8x8_ssse3
int dst_pitch
)
{
DECLARE_ALIGNED_ARRAY(16, unsigned char, FData2, 256);
DECLARE_ALIGNED(16, unsigned char, FData2[256]);
if (xoffset)
{
@ -528,7 +528,7 @@ void vp8_sixtap_predict8x4_ssse3
int dst_pitch
)
{
DECLARE_ALIGNED_ARRAY(16, unsigned char, FData2, 256);
DECLARE_ALIGNED(16, unsigned char, FData2[256]);
if (xoffset)
{
@ -576,7 +576,7 @@ void vp8_sixtap_predict4x4_ssse3
int dst_pitch
)
{
DECLARE_ALIGNED_ARRAY(16, unsigned char, FData2, 4*9);
DECLARE_ALIGNED(16, unsigned char, FData2[4*9]);
if (xoffset)
{

353
media/libvpx/vp8/common/x86/vp8_variance_impl_mmx.asm Normal file
Просмотреть файл

@ -0,0 +1,353 @@
;
; Copyright (c) 2010 The WebM project authors. All Rights Reserved.
;
; Use of this source code is governed by a BSD-style license
; that can be found in the LICENSE file in the root of the source
; tree. An additional intellectual property rights grant can be found
; in the file PATENTS. All contributing project authors may
; be found in the AUTHORS file in the root of the source tree.
;
%include "vpx_ports/x86_abi_support.asm"
%define mmx_filter_shift 7
;void vp8_filter_block2d_bil4x4_var_mmx
;(
; unsigned char *ref_ptr,
; int ref_pixels_per_line,
; unsigned char *src_ptr,
; int src_pixels_per_line,
; unsigned short *HFilter,
; unsigned short *VFilter,
; int *sum,
; unsigned int *sumsquared
;)
global sym(vp8_filter_block2d_bil4x4_var_mmx) PRIVATE
sym(vp8_filter_block2d_bil4x4_var_mmx):
push rbp
mov rbp, rsp
SHADOW_ARGS_TO_STACK 8
GET_GOT rbx
push rsi
push rdi
sub rsp, 16
; end prolog
pxor mm6, mm6 ;
pxor mm7, mm7 ;
mov rax, arg(4) ;HFilter ;
mov rdx, arg(5) ;VFilter ;
mov rsi, arg(0) ;ref_ptr ;
mov rdi, arg(2) ;src_ptr ;
mov rcx, 4 ;
pxor mm0, mm0 ;
movd mm1, [rsi] ;
movd mm3, [rsi+1] ;
punpcklbw mm1, mm0 ;
pmullw mm1, [rax] ;
punpcklbw mm3, mm0 ;
pmullw mm3, [rax+8] ;
paddw mm1, mm3 ;
paddw mm1, [GLOBAL(mmx_bi_rd)] ;
psraw mm1, mmx_filter_shift ;
movq mm5, mm1
%if ABI_IS_32BIT
add rsi, dword ptr arg(1) ;ref_pixels_per_line ;
%else
movsxd r8, dword ptr arg(1) ;ref_pixels_per_line ;
add rsi, r8
%endif
.filter_block2d_bil4x4_var_mmx_loop:
movd mm1, [rsi] ;
movd mm3, [rsi+1] ;
punpcklbw mm1, mm0 ;
pmullw mm1, [rax] ;
punpcklbw mm3, mm0 ;
pmullw mm3, [rax+8] ;
paddw mm1, mm3 ;
paddw mm1, [GLOBAL(mmx_bi_rd)] ;
psraw mm1, mmx_filter_shift ;
movq mm3, mm5 ;
movq mm5, mm1 ;
pmullw mm3, [rdx] ;
pmullw mm1, [rdx+8] ;
paddw mm1, mm3 ;
paddw mm1, [GLOBAL(mmx_bi_rd)] ;
psraw mm1, mmx_filter_shift ;
movd mm3, [rdi] ;
punpcklbw mm3, mm0 ;
psubw mm1, mm3 ;
paddw mm6, mm1 ;
pmaddwd mm1, mm1 ;
paddd mm7, mm1 ;
%if ABI_IS_32BIT
add rsi, dword ptr arg(1) ;ref_pixels_per_line ;
add rdi, dword ptr arg(3) ;src_pixels_per_line ;
%else
movsxd r8, dword ptr arg(1) ;ref_pixels_per_line
movsxd r9, dword ptr arg(3) ;src_pixels_per_line
add rsi, r8
add rdi, r9
%endif
sub rcx, 1 ;
jnz .filter_block2d_bil4x4_var_mmx_loop ;
pxor mm3, mm3 ;
pxor mm2, mm2 ;
punpcklwd mm2, mm6 ;
punpckhwd mm3, mm6 ;
paddd mm2, mm3 ;
movq mm6, mm2 ;
psrlq mm6, 32 ;
paddd mm2, mm6 ;
psrad mm2, 16 ;
movq mm4, mm7 ;
psrlq mm4, 32 ;
paddd mm4, mm7 ;
mov rdi, arg(6) ;sum
mov rsi, arg(7) ;sumsquared
movd dword ptr [rdi], mm2 ;
movd dword ptr [rsi], mm4 ;
; begin epilog
add rsp, 16
pop rdi
pop rsi
RESTORE_GOT
UNSHADOW_ARGS
pop rbp
ret
;void vp8_filter_block2d_bil_var_mmx
;(
; unsigned char *ref_ptr,
; int ref_pixels_per_line,
; unsigned char *src_ptr,
; int src_pixels_per_line,
; unsigned int Height,
; unsigned short *HFilter,
; unsigned short *VFilter,
; int *sum,
; unsigned int *sumsquared
;)
global sym(vp8_filter_block2d_bil_var_mmx) PRIVATE
sym(vp8_filter_block2d_bil_var_mmx):
push rbp
mov rbp, rsp
SHADOW_ARGS_TO_STACK 9
GET_GOT rbx
push rsi
push rdi
sub rsp, 16
; end prolog
pxor mm6, mm6 ;
pxor mm7, mm7 ;
mov rax, arg(5) ;HFilter ;
mov rdx, arg(6) ;VFilter ;
mov rsi, arg(0) ;ref_ptr ;
mov rdi, arg(2) ;src_ptr ;
movsxd rcx, dword ptr arg(4) ;Height ;
pxor mm0, mm0 ;
movq mm1, [rsi] ;
movq mm3, [rsi+1] ;
movq mm2, mm1 ;
movq mm4, mm3 ;
punpcklbw mm1, mm0 ;
punpckhbw mm2, mm0 ;
pmullw mm1, [rax] ;
pmullw mm2, [rax] ;
punpcklbw mm3, mm0 ;
punpckhbw mm4, mm0 ;
pmullw mm3, [rax+8] ;
pmullw mm4, [rax+8] ;
paddw mm1, mm3 ;
paddw mm2, mm4 ;
paddw mm1, [GLOBAL(mmx_bi_rd)] ;
psraw mm1, mmx_filter_shift ;
paddw mm2, [GLOBAL(mmx_bi_rd)] ;
psraw mm2, mmx_filter_shift ;
movq mm5, mm1
packuswb mm5, mm2 ;
%if ABI_IS_32BIT
add rsi, dword ptr arg(1) ;ref_pixels_per_line
%else
movsxd r8, dword ptr arg(1) ;ref_pixels_per_line
add rsi, r8
%endif
.filter_block2d_bil_var_mmx_loop:
movq mm1, [rsi] ;
movq mm3, [rsi+1] ;
movq mm2, mm1 ;
movq mm4, mm3 ;
punpcklbw mm1, mm0 ;
punpckhbw mm2, mm0 ;
pmullw mm1, [rax] ;
pmullw mm2, [rax] ;
punpcklbw mm3, mm0 ;
punpckhbw mm4, mm0 ;
pmullw mm3, [rax+8] ;
pmullw mm4, [rax+8] ;
paddw mm1, mm3 ;
paddw mm2, mm4 ;
paddw mm1, [GLOBAL(mmx_bi_rd)] ;
psraw mm1, mmx_filter_shift ;
paddw mm2, [GLOBAL(mmx_bi_rd)] ;
psraw mm2, mmx_filter_shift ;
movq mm3, mm5 ;
movq mm4, mm5 ;
punpcklbw mm3, mm0 ;
punpckhbw mm4, mm0 ;
movq mm5, mm1 ;
packuswb mm5, mm2 ;
pmullw mm3, [rdx] ;
pmullw mm4, [rdx] ;
pmullw mm1, [rdx+8] ;
pmullw mm2, [rdx+8] ;
paddw mm1, mm3 ;
paddw mm2, mm4 ;
paddw mm1, [GLOBAL(mmx_bi_rd)] ;
paddw mm2, [GLOBAL(mmx_bi_rd)] ;
psraw mm1, mmx_filter_shift ;
psraw mm2, mmx_filter_shift ;
movq mm3, [rdi] ;
movq mm4, mm3 ;
punpcklbw mm3, mm0 ;
punpckhbw mm4, mm0 ;
psubw mm1, mm3 ;
psubw mm2, mm4 ;
paddw mm6, mm1 ;
pmaddwd mm1, mm1 ;
paddw mm6, mm2 ;
pmaddwd mm2, mm2 ;
paddd mm7, mm1 ;
paddd mm7, mm2 ;
%if ABI_IS_32BIT
add rsi, dword ptr arg(1) ;ref_pixels_per_line ;
add rdi, dword ptr arg(3) ;src_pixels_per_line ;
%else
movsxd r8, dword ptr arg(1) ;ref_pixels_per_line ;
movsxd r9, dword ptr arg(3) ;src_pixels_per_line ;
add rsi, r8
add rdi, r9
%endif
sub rcx, 1 ;
jnz .filter_block2d_bil_var_mmx_loop ;
pxor mm3, mm3 ;
pxor mm2, mm2 ;
punpcklwd mm2, mm6 ;
punpckhwd mm3, mm6 ;
paddd mm2, mm3 ;
movq mm6, mm2 ;
psrlq mm6, 32 ;
paddd mm2, mm6 ;
psrad mm2, 16 ;
movq mm4, mm7 ;
psrlq mm4, 32 ;
paddd mm4, mm7 ;
mov rdi, arg(7) ;sum
mov rsi, arg(8) ;sumsquared
movd dword ptr [rdi], mm2 ;
movd dword ptr [rsi], mm4 ;
; begin epilog
add rsp, 16
pop rdi
pop rsi
RESTORE_GOT
UNSHADOW_ARGS
pop rbp
ret
SECTION_RODATA
;short mmx_bi_rd[4] = { 64, 64, 64, 64};
align 16
mmx_bi_rd:
times 4 dw 64

155
media/libvpx/vp8/common/x86/variance_mmx.c → media/libvpx/vp8/common/x86/vp8_variance_mmx.c
Просмотреть файл

@ -8,6 +8,7 @@
* be found in the AUTHORS file in the root of the source tree.
*/
#include "./vp8_rtcd.h"
#include "vpx_config.h"
#include "vp8/common/variance.h"
#include "vpx_ports/mem.h"
@ -34,25 +35,6 @@ extern void filter_block1d_v6_mmx
short *filter
);
extern unsigned int vp8_get_mb_ss_mmx(const short *src_ptr);
extern unsigned int vp8_get8x8var_mmx
(
const unsigned char *src_ptr,
int source_stride,
const unsigned char *ref_ptr,
int recon_stride,
unsigned int *SSE,
int *Sum
);
extern unsigned int vp8_get4x4var_mmx
(
const unsigned char *src_ptr,
int source_stride,
const unsigned char *ref_ptr,
int recon_stride,
unsigned int *SSE,
int *Sum
);
extern void vp8_filter_block2d_bil4x4_var_mmx
(
const unsigned char *ref_ptr,
@ -77,127 +59,6 @@ extern void vp8_filter_block2d_bil_var_mmx
unsigned int *sumsquared
);
unsigned int vp8_variance4x4_mmx(
const unsigned char *src_ptr,
int source_stride,
const unsigned char *ref_ptr,
int recon_stride,
unsigned int *sse)
{
unsigned int var;
int avg;
vp8_get4x4var_mmx(src_ptr, source_stride, ref_ptr, recon_stride, &var, &avg) ;
*sse = var;
return (var - (((unsigned int)avg * avg) >> 4));
}
unsigned int vp8_variance8x8_mmx(
const unsigned char *src_ptr,
int source_stride,
const unsigned char *ref_ptr,
int recon_stride,
unsigned int *sse)
{
unsigned int var;
int avg;
vp8_get8x8var_mmx(src_ptr, source_stride, ref_ptr, recon_stride, &var, &avg) ;
*sse = var;
return (var - (((unsigned int)avg * avg) >> 6));
}
unsigned int vp8_mse16x16_mmx(
const unsigned char *src_ptr,
int source_stride,
const unsigned char *ref_ptr,
int recon_stride,
unsigned int *sse)
{
unsigned int sse0, sse1, sse2, sse3, var;
int sum0, sum1, sum2, sum3;
vp8_get8x8var_mmx(src_ptr, source_stride, ref_ptr, recon_stride, &sse0, &sum0) ;
vp8_get8x8var_mmx(src_ptr + 8, source_stride, ref_ptr + 8, recon_stride, &sse1, &sum1);
vp8_get8x8var_mmx(src_ptr + 8 * source_stride, source_stride, ref_ptr + 8 * recon_stride, recon_stride, &sse2, &sum2) ;
vp8_get8x8var_mmx(src_ptr + 8 * source_stride + 8, source_stride, ref_ptr + 8 * recon_stride + 8, recon_stride, &sse3, &sum3);
var = sse0 + sse1 + sse2 + sse3;
*sse = var;
return var;
}
unsigned int vp8_variance16x16_mmx(
const unsigned char *src_ptr,
int source_stride,
const unsigned char *ref_ptr,
int recon_stride,
unsigned int *sse)
{
unsigned int sse0, sse1, sse2, sse3, var;
int sum0, sum1, sum2, sum3, avg;
vp8_get8x8var_mmx(src_ptr, source_stride, ref_ptr, recon_stride, &sse0, &sum0) ;
vp8_get8x8var_mmx(src_ptr + 8, source_stride, ref_ptr + 8, recon_stride, &sse1, &sum1);
vp8_get8x8var_mmx(src_ptr + 8 * source_stride, source_stride, ref_ptr + 8 * recon_stride, recon_stride, &sse2, &sum2) ;
vp8_get8x8var_mmx(src_ptr + 8 * source_stride + 8, source_stride, ref_ptr + 8 * recon_stride + 8, recon_stride, &sse3, &sum3);
var = sse0 + sse1 + sse2 + sse3;
avg = sum0 + sum1 + sum2 + sum3;
*sse = var;
return (var - (((unsigned int)avg * avg) >> 8));
}
unsigned int vp8_variance16x8_mmx(
const unsigned char *src_ptr,
int source_stride,
const unsigned char *ref_ptr,
int recon_stride,
unsigned int *sse)
{
unsigned int sse0, sse1, var;
int sum0, sum1, avg;
vp8_get8x8var_mmx(src_ptr, source_stride, ref_ptr, recon_stride, &sse0, &sum0) ;
vp8_get8x8var_mmx(src_ptr + 8, source_stride, ref_ptr + 8, recon_stride, &sse1, &sum1);
var = sse0 + sse1;
avg = sum0 + sum1;
*sse = var;
return (var - (((unsigned int)avg * avg) >> 7));
}
unsigned int vp8_variance8x16_mmx(
const unsigned char *src_ptr,
int source_stride,
const unsigned char *ref_ptr,
int recon_stride,
unsigned int *sse)
{
unsigned int sse0, sse1, var;
int sum0, sum1, avg;
vp8_get8x8var_mmx(src_ptr, source_stride, ref_ptr, recon_stride, &sse0, &sum0) ;
vp8_get8x8var_mmx(src_ptr + 8 * source_stride, source_stride, ref_ptr + 8 * recon_stride, recon_stride, &sse1, &sum1) ;
var = sse0 + sse1;
avg = sum0 + sum1;
*sse = var;
return (var - (((unsigned int)avg * avg) >> 7));
}
unsigned int vp8_sub_pixel_variance4x4_mmx
(
const unsigned char *src_ptr,
@ -286,20 +147,6 @@ unsigned int vp8_sub_pixel_variance16x16_mmx
}
unsigned int vp8_sub_pixel_mse16x16_mmx(
const unsigned char *src_ptr,
int src_pixels_per_line,
int xoffset,
int yoffset,
const unsigned char *dst_ptr,
int dst_pixels_per_line,
unsigned int *sse
)
{
vp8_sub_pixel_variance16x16_mmx(src_ptr, src_pixels_per_line, xoffset, yoffset, dst_ptr, dst_pixels_per_line, sse);
return *sse;
}
unsigned int vp8_sub_pixel_variance16x8_mmx
(
const unsigned char *src_ptr,

156
media/libvpx/vp8/common/x86/variance_sse2.c → media/libvpx/vp8/common/x86/vp8_variance_sse2.c
Просмотреть файл

@ -8,6 +8,7 @@
* be found in the AUTHORS file in the root of the source tree.
*/
#include "./vp8_rtcd.h"
#include "vpx_config.h"
#include "vp8/common/variance.h"
#include "vpx_ports/mem.h"
@ -30,38 +31,6 @@ extern void vp8_filter_block2d_bil4x4_var_mmx
unsigned int *sumsquared
);
extern unsigned int vp8_get4x4var_mmx
(
const unsigned char *src_ptr,
int source_stride,
const unsigned char *ref_ptr,
int recon_stride,
unsigned int *SSE,
int *Sum
);
unsigned int vp8_get_mb_ss_sse2
(
const short *src_ptr
);
unsigned int vp8_get16x16var_sse2
(
const unsigned char *src_ptr,
int source_stride,
const unsigned char *ref_ptr,
int recon_stride,
unsigned int *SSE,
int *Sum
);
unsigned int vp8_get8x8var_sse2
(
const unsigned char *src_ptr,
int source_stride,
const unsigned char *ref_ptr,
int recon_stride,
unsigned int *SSE,
int *Sum
);
void vp8_filter_block2d_bil_var_sse2
(
const unsigned char *ref_ptr,
@ -135,115 +104,6 @@ void vp8_half_vert_variance16x_h_sse2
unsigned int *sumsquared
);
unsigned int vp8_variance4x4_wmt(
const unsigned char *src_ptr,
int source_stride,
const unsigned char *ref_ptr,
int recon_stride,
unsigned int *sse)
{
unsigned int var;
int avg;
vp8_get4x4var_mmx(src_ptr, source_stride, ref_ptr, recon_stride, &var, &avg) ;
*sse = var;
return (var - (((unsigned int)avg * avg) >> 4));
}
unsigned int vp8_variance8x8_wmt
(
const unsigned char *src_ptr,
int source_stride,
const unsigned char *ref_ptr,
int recon_stride,
unsigned int *sse)
{
unsigned int var;
int avg;
vp8_get8x8var_sse2(src_ptr, source_stride, ref_ptr, recon_stride, &var, &avg) ;
*sse = var;
return (var - (((unsigned int)avg * avg) >> 6));
}
unsigned int vp8_variance16x16_wmt
(
const unsigned char *src_ptr,
int source_stride,
const unsigned char *ref_ptr,
int recon_stride,
unsigned int *sse)
{
unsigned int sse0;
int sum0;
vp8_get16x16var_sse2(src_ptr, source_stride, ref_ptr, recon_stride, &sse0, &sum0) ;
*sse = sse0;
return (sse0 - (((unsigned int)sum0 * sum0) >> 8));
}
unsigned int vp8_mse16x16_wmt(
const unsigned char *src_ptr,
int source_stride,
const unsigned char *ref_ptr,
int recon_stride,
unsigned int *sse)
{
unsigned int sse0;
int sum0;
vp8_get16x16var_sse2(src_ptr, source_stride, ref_ptr, recon_stride, &sse0, &sum0) ;
*sse = sse0;
return sse0;
}
unsigned int vp8_variance16x8_wmt
(
const unsigned char *src_ptr,
int source_stride,
const unsigned char *ref_ptr,
int recon_stride,
unsigned int *sse)
{
unsigned int sse0, sse1, var;
int sum0, sum1, avg;
vp8_get8x8var_sse2(src_ptr, source_stride, ref_ptr, recon_stride, &sse0, &sum0) ;
vp8_get8x8var_sse2(src_ptr + 8, source_stride, ref_ptr + 8, recon_stride, &sse1, &sum1);
var = sse0 + sse1;
avg = sum0 + sum1;
*sse = var;
return (var - (((unsigned int)avg * avg) >> 7));
}
unsigned int vp8_variance8x16_wmt
(
const unsigned char *src_ptr,
int source_stride,
const unsigned char *ref_ptr,
int recon_stride,
unsigned int *sse)
{
unsigned int sse0, sse1, var;
int sum0, sum1, avg;
vp8_get8x8var_sse2(src_ptr, source_stride, ref_ptr, recon_stride, &sse0, &sum0) ;
vp8_get8x8var_sse2(src_ptr + 8 * source_stride, source_stride, ref_ptr + 8 * recon_stride, recon_stride, &sse1, &sum1) ;
var = sse0 + sse1;
avg = sum0 + sum1;
*sse = var;
return (var - (((unsigned int)avg * avg) >> 7));
}
unsigned int vp8_sub_pixel_variance4x4_wmt
(
const unsigned char *src_ptr,
@ -378,20 +238,6 @@ unsigned int vp8_sub_pixel_variance16x16_wmt
return (xxsum0 - (((unsigned int)xsum0 * xsum0) >> 8));
}
unsigned int vp8_sub_pixel_mse16x16_wmt(
const unsigned char *src_ptr,
int src_pixels_per_line,
int xoffset,
int yoffset,
const unsigned char *dst_ptr,
int dst_pixels_per_line,
unsigned int *sse
)
{
vp8_sub_pixel_variance16x16_wmt(src_ptr, src_pixels_per_line, xoffset, yoffset, dst_ptr, dst_pixels_per_line, sse);
return *sse;
}
unsigned int vp8_sub_pixel_variance16x8_wmt
(
const unsigned char *src_ptr,

63
media/libvpx/vp8/decoder/decodeframe.c
Просмотреть файл

@ -142,7 +142,7 @@ static void decode_macroblock(VP8D_COMP *pbi, MACROBLOCKD *xd,
* Better to use the predictor as reconstruction.
*/
pbi->frame_corrupt_residual = 1;
vpx_memset(xd->qcoeff, 0, sizeof(xd->qcoeff));
memset(xd->qcoeff, 0, sizeof(xd->qcoeff));
vp8_conceal_corrupt_mb(xd);
@ -151,7 +151,7 @@ static void decode_macroblock(VP8D_COMP *pbi, MACROBLOCKD *xd,
/* force idct to be skipped for B_PRED and use the
* prediction only for reconstruction
* */
vpx_memset(xd->eobs, 0, 25);
memset(xd->eobs, 0, 25);
}
}
#endif
@ -184,7 +184,7 @@ static void decode_macroblock(VP8D_COMP *pbi, MACROBLOCKD *xd,
/* clear out residual eob info */
if(xd->mode_info_context->mbmi.mb_skip_coeff)
vpx_memset(xd->eobs, 0, 25);
memset(xd->eobs, 0, 25);
intra_prediction_down_copy(xd, xd->recon_above[0] + 16);
@ -214,7 +214,7 @@ static void decode_macroblock(VP8D_COMP *pbi, MACROBLOCKD *xd,
(b->qcoeff[0] * DQC[0],
dst, dst_stride,
dst, dst_stride);
vpx_memset(b->qcoeff, 0, 2 * sizeof(b->qcoeff[0]));
memset(b->qcoeff, 0, 2 * sizeof(b->qcoeff[0]));
}
}
}
@ -251,14 +251,14 @@ static void decode_macroblock(VP8D_COMP *pbi, MACROBLOCKD *xd,
vp8_short_inv_walsh4x4(&b->dqcoeff[0],
xd->qcoeff);
vpx_memset(b->qcoeff, 0, 16 * sizeof(b->qcoeff[0]));
memset(b->qcoeff, 0, 16 * sizeof(b->qcoeff[0]));
}
else
{
b->dqcoeff[0] = b->qcoeff[0] * xd->dequant_y2[0];
vp8_short_inv_walsh4x4_1(&b->dqcoeff[0],
xd->qcoeff);
vpx_memset(b->qcoeff, 0, 2 * sizeof(b->qcoeff[0]));
memset(b->qcoeff, 0, 2 * sizeof(b->qcoeff[0]));
}
/* override the dc dequant constant in order to preserve the
@ -323,7 +323,7 @@ static void yv12_extend_frame_top_c(YV12_BUFFER_CONFIG *ybf)
for (i = 0; i < (int)Border; i++)
{
vpx_memcpy(dest_ptr1, src_ptr1, plane_stride);
memcpy(dest_ptr1, src_ptr1, plane_stride);
dest_ptr1 += plane_stride;
}
@ -338,7 +338,7 @@ static void yv12_extend_frame_top_c(YV12_BUFFER_CONFIG *ybf)
for (i = 0; i < (int)(Border); i++)
{
vpx_memcpy(dest_ptr1, src_ptr1, plane_stride);
memcpy(dest_ptr1, src_ptr1, plane_stride);
dest_ptr1 += plane_stride;
}
@ -351,7 +351,7 @@ static void yv12_extend_frame_top_c(YV12_BUFFER_CONFIG *ybf)
for (i = 0; i < (int)(Border); i++)
{
vpx_memcpy(dest_ptr1, src_ptr1, plane_stride);
memcpy(dest_ptr1, src_ptr1, plane_stride);
dest_ptr1 += plane_stride;
}
}
@ -379,7 +379,7 @@ static void yv12_extend_frame_bottom_c(YV12_BUFFER_CONFIG *ybf)
for (i = 0; i < (int)Border; i++)
{
vpx_memcpy(dest_ptr2, src_ptr2, plane_stride);
memcpy(dest_ptr2, src_ptr2, plane_stride);
dest_ptr2 += plane_stride;
}
@ -397,7 +397,7 @@ static void yv12_extend_frame_bottom_c(YV12_BUFFER_CONFIG *ybf)
for (i = 0; i < (int)(Border); i++)
{
vpx_memcpy(dest_ptr2, src_ptr2, plane_stride);
memcpy(dest_ptr2, src_ptr2, plane_stride);
dest_ptr2 += plane_stride;
}
@ -411,7 +411,7 @@ static void yv12_extend_frame_bottom_c(YV12_BUFFER_CONFIG *ybf)
for (i = 0; i < (int)(Border); i++)
{
vpx_memcpy(dest_ptr2, src_ptr2, plane_stride);
memcpy(dest_ptr2, src_ptr2, plane_stride);
dest_ptr2 += plane_stride;
}
}
@ -446,8 +446,8 @@ static void yv12_extend_frame_left_right_c(YV12_BUFFER_CONFIG *ybf,
for (i = 0; i < plane_height; i++)
{
vpx_memset(dest_ptr1, src_ptr1[0], Border);
vpx_memset(dest_ptr2, src_ptr2[0], Border);
memset(dest_ptr1, src_ptr1[0], Border);
memset(dest_ptr2, src_ptr2[0], Border);
src_ptr1 += plane_stride;
src_ptr2 += plane_stride;
dest_ptr1 += plane_stride;
@ -470,8 +470,8 @@ static void yv12_extend_frame_left_right_c(YV12_BUFFER_CONFIG *ybf,
for (i = 0; i < plane_height; i++)
{
vpx_memset(dest_ptr1, src_ptr1[0], Border);
vpx_memset(dest_ptr2, src_ptr2[0], Border);
memset(dest_ptr1, src_ptr1[0], Border);
memset(dest_ptr2, src_ptr2[0], Border);
src_ptr1 += plane_stride;
src_ptr2 += plane_stride;
dest_ptr1 += plane_stride;
@ -490,8 +490,8 @@ static void yv12_extend_frame_left_right_c(YV12_BUFFER_CONFIG *ybf,
for (i = 0; i < plane_height; i++)
{
vpx_memset(dest_ptr1, src_ptr1[0], Border);
vpx_memset(dest_ptr2, src_ptr2[0], Border);
memset(dest_ptr1, src_ptr1[0], Border);
memset(dest_ptr2, src_ptr2[0], Border);
src_ptr1 += plane_stride;
src_ptr2 += plane_stride;
dest_ptr1 += plane_stride;
@ -568,7 +568,7 @@ static void decode_mb_rows(VP8D_COMP *pbi)
/* reset contexts */
xd->above_context = pc->above_context;
vpx_memset(xd->left_context, 0, sizeof(ENTROPY_CONTEXT_PLANES));
memset(xd->left_context, 0, sizeof(ENTROPY_CONTEXT_PLANES));
xd->left_available = 0;
@ -918,19 +918,19 @@ static void init_frame(VP8D_COMP *pbi)
if (pc->frame_type == KEY_FRAME)
{
/* Various keyframe initializations */
vpx_memcpy(pc->fc.mvc, vp8_default_mv_context, sizeof(vp8_default_mv_context));
memcpy(pc->fc.mvc, vp8_default_mv_context, sizeof(vp8_default_mv_context));
vp8_init_mbmode_probs(pc);
vp8_default_coef_probs(pc);
/* reset the segment feature data to 0 with delta coding (Default state). */
vpx_memset(xd->segment_feature_data, 0, sizeof(xd->segment_feature_data));
memset(xd->segment_feature_data, 0, sizeof(xd->segment_feature_data));
xd->mb_segement_abs_delta = SEGMENT_DELTADATA;
/* reset the mode ref deltasa for loop filter */
vpx_memset(xd->ref_lf_deltas, 0, sizeof(xd->ref_lf_deltas));
vpx_memset(xd->mode_lf_deltas, 0, sizeof(xd->mode_lf_deltas));
memset(xd->ref_lf_deltas, 0, sizeof(xd->ref_lf_deltas));
memset(xd->mode_lf_deltas, 0, sizeof(xd->mode_lf_deltas));
/* All buffers are implicitly updated on key frames. */
pc->refresh_golden_frame = 1;
@ -1069,12 +1069,11 @@ int vp8_decode_frame(VP8D_COMP *pbi)
pc->vert_scale = clear[6] >> 6;
}
data += 7;
clear += 7;
}
else
{
vpx_memcpy(&xd->pre, yv12_fb_new, sizeof(YV12_BUFFER_CONFIG));
vpx_memcpy(&xd->dst, yv12_fb_new, sizeof(YV12_BUFFER_CONFIG));
memcpy(&xd->pre, yv12_fb_new, sizeof(YV12_BUFFER_CONFIG));
memcpy(&xd->dst, yv12_fb_new, sizeof(YV12_BUFFER_CONFIG));
}
}
if ((!pbi->decoded_key_frame && pc->frame_type != KEY_FRAME))
@ -1106,7 +1105,7 @@ int vp8_decode_frame(VP8D_COMP *pbi)
{
xd->mb_segement_abs_delta = (unsigned char)vp8_read_bit(bc);
vpx_memset(xd->segment_feature_data, 0, sizeof(xd->segment_feature_data));
memset(xd->segment_feature_data, 0, sizeof(xd->segment_feature_data));
/* For each segmentation feature (Quant and loop filter level) */
for (i = 0; i < MB_LVL_MAX; i++)
@ -1130,7 +1129,7 @@ int vp8_decode_frame(VP8D_COMP *pbi)
if (xd->update_mb_segmentation_map)
{
/* Which macro block level features are enabled */
vpx_memset(xd->mb_segment_tree_probs, 255, sizeof(xd->mb_segment_tree_probs));
memset(xd->mb_segment_tree_probs, 255, sizeof(xd->mb_segment_tree_probs));
/* Read the probs used to decode the segment id for each macro block. */
for (i = 0; i < MB_FEATURE_TREE_PROBS; i++)
@ -1279,7 +1278,7 @@ int vp8_decode_frame(VP8D_COMP *pbi)
#endif
if (pc->refresh_entropy_probs == 0)
{
vpx_memcpy(&pc->lfc, &pc->fc, sizeof(pc->fc));
memcpy(&pc->lfc, &pc->fc, sizeof(pc->fc));
}
pc->refresh_last_frame = pc->frame_type == KEY_FRAME || vp8_read_bit(bc);
@ -1328,7 +1327,7 @@ int vp8_decode_frame(VP8D_COMP *pbi)
}
/* clear out the coeff buffer */
vpx_memset(xd->qcoeff, 0, sizeof(xd->qcoeff));
memset(xd->qcoeff, 0, sizeof(xd->qcoeff));
vp8_decode_mode_mvs(pbi);
@ -1342,7 +1341,7 @@ int vp8_decode_frame(VP8D_COMP *pbi)
}
#endif
vpx_memset(pc->above_context, 0, sizeof(ENTROPY_CONTEXT_PLANES) * pc->mb_cols);
memset(pc->above_context, 0, sizeof(ENTROPY_CONTEXT_PLANES) * pc->mb_cols);
pbi->frame_corrupt_residual = 0;
#if CONFIG_MULTITHREAD
@ -1381,7 +1380,7 @@ int vp8_decode_frame(VP8D_COMP *pbi)
if (pc->refresh_entropy_probs == 0)
{
vpx_memcpy(&pc->fc, &pc->lfc, sizeof(pc->fc));
memcpy(&pc->fc, &pc->lfc, sizeof(pc->fc));
pbi->independent_partitions = prev_independent_partitions;
}

4
media/libvpx/vp8/decoder/detokenize.c
Просмотреть файл

@ -20,8 +20,8 @@ void vp8_reset_mb_tokens_context(MACROBLOCKD *x)
ENTROPY_CONTEXT *a_ctx = ((ENTROPY_CONTEXT *)x->above_context);
ENTROPY_CONTEXT *l_ctx = ((ENTROPY_CONTEXT *)x->left_context);
vpx_memset(a_ctx, 0, sizeof(ENTROPY_CONTEXT_PLANES)-1);
vpx_memset(l_ctx, 0, sizeof(ENTROPY_CONTEXT_PLANES)-1);
memset(a_ctx, 0, sizeof(ENTROPY_CONTEXT_PLANES)-1);
memset(l_ctx, 0, sizeof(ENTROPY_CONTEXT_PLANES)-1);
/* Clear entropy contexts for Y2 blocks */
if (!x->mode_info_context->mbmi.is_4x4)

2
media/libvpx/vp8/decoder/error_concealment.c
Просмотреть файл

@ -350,7 +350,7 @@ static void estimate_missing_mvs(MB_OVERLAP *overlaps,
unsigned int first_corrupt)
{
int mb_row, mb_col;
vpx_memset(overlaps, 0, sizeof(MB_OVERLAP) * mb_rows * mb_cols);
memset(overlaps, 0, sizeof(MB_OVERLAP) * mb_rows * mb_cols);
/* First calculate the overlaps for all blocks */
for (mb_row = 0; mb_row < mb_rows; ++mb_row)
{

4
media/libvpx/vp8/decoder/onyxd_if.c
Просмотреть файл

@ -58,7 +58,7 @@ static struct VP8D_COMP * create_decompressor(VP8D_CONFIG *oxcf)
if (!pbi)
return NULL;
vpx_memset(pbi, 0, sizeof(VP8D_COMP));
memset(pbi, 0, sizeof(VP8D_COMP));
if (setjmp(pbi->common.error.jmp))
{
@ -259,7 +259,7 @@ static int swap_frame_buffers (VP8_COMMON *cm)
return err;
}
int check_fragments_for_errors(VP8D_COMP *pbi)
static int check_fragments_for_errors(VP8D_COMP *pbi)
{
if (!pbi->ec_active &&
pbi->fragments.count <= 1 && pbi->fragments.sizes[0] == 0)

52
media/libvpx/vp8/decoder/threading.c
Просмотреть файл

@ -60,12 +60,12 @@ static void setup_decoding_thread_data(VP8D_COMP *pbi, MACROBLOCKD *xd, MB_ROW_D
mbd->segmentation_enabled = xd->segmentation_enabled;
mbd->mb_segement_abs_delta = xd->mb_segement_abs_delta;
vpx_memcpy(mbd->segment_feature_data, xd->segment_feature_data, sizeof(xd->segment_feature_data));
memcpy(mbd->segment_feature_data, xd->segment_feature_data, sizeof(xd->segment_feature_data));
/*signed char ref_lf_deltas[MAX_REF_LF_DELTAS];*/
vpx_memcpy(mbd->ref_lf_deltas, xd->ref_lf_deltas, sizeof(xd->ref_lf_deltas));
memcpy(mbd->ref_lf_deltas, xd->ref_lf_deltas, sizeof(xd->ref_lf_deltas));
/*signed char mode_lf_deltas[MAX_MODE_LF_DELTAS];*/
vpx_memcpy(mbd->mode_lf_deltas, xd->mode_lf_deltas, sizeof(xd->mode_lf_deltas));
memcpy(mbd->mode_lf_deltas, xd->mode_lf_deltas, sizeof(xd->mode_lf_deltas));
/*unsigned char mode_ref_lf_delta_enabled;
unsigned char mode_ref_lf_delta_update;*/
mbd->mode_ref_lf_delta_enabled = xd->mode_ref_lf_delta_enabled;
@ -73,10 +73,10 @@ static void setup_decoding_thread_data(VP8D_COMP *pbi, MACROBLOCKD *xd, MB_ROW_D
mbd->current_bc = &pbi->mbc[0];
vpx_memcpy(mbd->dequant_y1_dc, xd->dequant_y1_dc, sizeof(xd->dequant_y1_dc));
vpx_memcpy(mbd->dequant_y1, xd->dequant_y1, sizeof(xd->dequant_y1));
vpx_memcpy(mbd->dequant_y2, xd->dequant_y2, sizeof(xd->dequant_y2));
vpx_memcpy(mbd->dequant_uv, xd->dequant_uv, sizeof(xd->dequant_uv));
memcpy(mbd->dequant_y1_dc, xd->dequant_y1_dc, sizeof(xd->dequant_y1_dc));
memcpy(mbd->dequant_y1, xd->dequant_y1, sizeof(xd->dequant_y1));
memcpy(mbd->dequant_y2, xd->dequant_y2, sizeof(xd->dequant_y2));
memcpy(mbd->dequant_uv, xd->dequant_uv, sizeof(xd->dequant_uv));
mbd->fullpixel_mask = 0xffffffff;
@ -137,7 +137,7 @@ static void mt_decode_macroblock(VP8D_COMP *pbi, MACROBLOCKD *xd,
* Better to use the predictor as reconstruction.
*/
pbi->frame_corrupt_residual = 1;
vpx_memset(xd->qcoeff, 0, sizeof(xd->qcoeff));
memset(xd->qcoeff, 0, sizeof(xd->qcoeff));
vp8_conceal_corrupt_mb(xd);
@ -146,7 +146,7 @@ static void mt_decode_macroblock(VP8D_COMP *pbi, MACROBLOCKD *xd,
/* force idct to be skipped for B_PRED and use the
* prediction only for reconstruction
* */
vpx_memset(xd->eobs, 0, 25);
memset(xd->eobs, 0, 25);
}
}
#endif
@ -179,7 +179,7 @@ static void mt_decode_macroblock(VP8D_COMP *pbi, MACROBLOCKD *xd,
/* clear out residual eob info */
if(xd->mode_info_context->mbmi.mb_skip_coeff)
vpx_memset(xd->eobs, 0, 25);
memset(xd->eobs, 0, 25);
intra_prediction_down_copy(xd, xd->recon_above[0] + 16);
@ -229,7 +229,7 @@ static void mt_decode_macroblock(VP8D_COMP *pbi, MACROBLOCKD *xd,
{
vp8_dc_only_idct_add(b->qcoeff[0] * DQC[0],
dst, dst_stride, dst, dst_stride);
vpx_memset(b->qcoeff, 0, 2 * sizeof(b->qcoeff[0]));
memset(b->qcoeff, 0, 2 * sizeof(b->qcoeff[0]));
}
}
}
@ -266,14 +266,14 @@ static void mt_decode_macroblock(VP8D_COMP *pbi, MACROBLOCKD *xd,
vp8_short_inv_walsh4x4(&b->dqcoeff[0],
xd->qcoeff);
vpx_memset(b->qcoeff, 0, 16 * sizeof(b->qcoeff[0]));
memset(b->qcoeff, 0, 16 * sizeof(b->qcoeff[0]));
}
else
{
b->dqcoeff[0] = b->qcoeff[0] * xd->dequant_y2[0];
vp8_short_inv_walsh4x4_1(&b->dqcoeff[0],
xd->qcoeff);
vpx_memset(b->qcoeff, 0, 2 * sizeof(b->qcoeff[0]));
memset(b->qcoeff, 0, 2 * sizeof(b->qcoeff[0]));
}
/* override the dc dequant constant in order to preserve the
@ -360,7 +360,7 @@ static void mt_decode_mb_rows(VP8D_COMP *pbi, MACROBLOCKD *xd, int start_mb_row)
/* reset contexts */
xd->above_context = pc->above_context;
vpx_memset(xd->left_context, 0, sizeof(ENTROPY_CONTEXT_PLANES));
memset(xd->left_context, 0, sizeof(ENTROPY_CONTEXT_PLANES));
xd->left_available = 0;
@ -499,9 +499,9 @@ static void mt_decode_mb_rows(VP8D_COMP *pbi, MACROBLOCKD *xd, int start_mb_row)
if( mb_row != pc->mb_rows-1 )
{
/* Save decoded MB last row data for next-row decoding */
vpx_memcpy((pbi->mt_yabove_row[mb_row + 1] + 32 + mb_col*16), (xd->dst.y_buffer + 15 * recon_y_stride), 16);
vpx_memcpy((pbi->mt_uabove_row[mb_row + 1] + 16 + mb_col*8), (xd->dst.u_buffer + 7 * recon_uv_stride), 8);
vpx_memcpy((pbi->mt_vabove_row[mb_row + 1] + 16 + mb_col*8), (xd->dst.v_buffer + 7 * recon_uv_stride), 8);
memcpy((pbi->mt_yabove_row[mb_row + 1] + 32 + mb_col*16), (xd->dst.y_buffer + 15 * recon_y_stride), 16);
memcpy((pbi->mt_uabove_row[mb_row + 1] + 16 + mb_col*8), (xd->dst.u_buffer + 7 * recon_uv_stride), 8);
memcpy((pbi->mt_vabove_row[mb_row + 1] + 16 + mb_col*8), (xd->dst.v_buffer + 7 * recon_uv_stride), 8);
}
/* save left_col for next MB decoding */
@ -876,23 +876,23 @@ void vp8mt_decode_mb_rows( VP8D_COMP *pbi, MACROBLOCKD *xd)
if (filter_level)
{
/* Set above_row buffer to 127 for decoding first MB row */
vpx_memset(pbi->mt_yabove_row[0] + VP8BORDERINPIXELS-1, 127, yv12_fb_new->y_width + 5);
vpx_memset(pbi->mt_uabove_row[0] + (VP8BORDERINPIXELS>>1)-1, 127, (yv12_fb_new->y_width>>1) +5);
vpx_memset(pbi->mt_vabove_row[0] + (VP8BORDERINPIXELS>>1)-1, 127, (yv12_fb_new->y_width>>1) +5);
memset(pbi->mt_yabove_row[0] + VP8BORDERINPIXELS-1, 127, yv12_fb_new->y_width + 5);
memset(pbi->mt_uabove_row[0] + (VP8BORDERINPIXELS>>1)-1, 127, (yv12_fb_new->y_width>>1) +5);
memset(pbi->mt_vabove_row[0] + (VP8BORDERINPIXELS>>1)-1, 127, (yv12_fb_new->y_width>>1) +5);
for (j=1; j<pc->mb_rows; j++)
{
vpx_memset(pbi->mt_yabove_row[j] + VP8BORDERINPIXELS-1, (unsigned char)129, 1);
vpx_memset(pbi->mt_uabove_row[j] + (VP8BORDERINPIXELS>>1)-1, (unsigned char)129, 1);
vpx_memset(pbi->mt_vabove_row[j] + (VP8BORDERINPIXELS>>1)-1, (unsigned char)129, 1);
memset(pbi->mt_yabove_row[j] + VP8BORDERINPIXELS-1, (unsigned char)129, 1);
memset(pbi->mt_uabove_row[j] + (VP8BORDERINPIXELS>>1)-1, (unsigned char)129, 1);
memset(pbi->mt_vabove_row[j] + (VP8BORDERINPIXELS>>1)-1, (unsigned char)129, 1);
}
/* Set left_col to 129 initially */
for (j=0; j<pc->mb_rows; j++)
{
vpx_memset(pbi->mt_yleft_col[j], (unsigned char)129, 16);
vpx_memset(pbi->mt_uleft_col[j], (unsigned char)129, 8);
vpx_memset(pbi->mt_vleft_col[j], (unsigned char)129, 8);
memset(pbi->mt_yleft_col[j], (unsigned char)129, 16);
memset(pbi->mt_uleft_col[j], (unsigned char)129, 8);
memset(pbi->mt_vleft_col[j], (unsigned char)129, 8);
}
/* Initialize the loop filter for this frame. */

138
media/libvpx/vp8/encoder/arm/armv6/vp8_mse16x16_armv6.asm
Просмотреть файл

@ -1,138 +0,0 @@
;
; Copyright (c) 2011 The WebM project authors. All Rights Reserved.
;
; Use of this source code is governed by a BSD-style license
; that can be found in the LICENSE file in the root of the source
; tree. An additional intellectual property rights grant can be found
; in the file PATENTS. All contributing project authors may
; be found in the AUTHORS file in the root of the source tree.
;
EXPORT |vp8_mse16x16_armv6|
ARM
AREA ||.text||, CODE, READONLY, ALIGN=2
; r0 unsigned char *src_ptr
; r1 int source_stride
; r2 unsigned char *ref_ptr
; r3 int recon_stride
; stack unsigned int *sse
;
;note: Based on vp8_variance16x16_armv6. In this function, sum is never used.
; So, we can remove this part of calculation.
|vp8_mse16x16_armv6| PROC
push {r4-r9, lr}
pld [r0, r1, lsl #0]
pld [r2, r3, lsl #0]
mov r12, #16 ; set loop counter to 16 (=block height)
mov r4, #0 ; initialize sse = 0
loop
; 1st 4 pixels
ldr r5, [r0, #0x0] ; load 4 src pixels
ldr r6, [r2, #0x0] ; load 4 ref pixels
mov lr, #0 ; constant zero
usub8 r8, r5, r6 ; calculate difference
pld [r0, r1, lsl #1]
sel r7, r8, lr ; select bytes with positive difference
usub8 r9, r6, r5 ; calculate difference with reversed operands
pld [r2, r3, lsl #1]
sel r8, r9, lr ; select bytes with negative difference
; calculate partial sums
usad8 r5, r7, lr ; calculate sum of positive differences
usad8 r6, r8, lr ; calculate sum of negative differences
orr r8, r8, r7 ; differences of all 4 pixels
ldr r5, [r0, #0x4] ; load 4 src pixels
; calculate sse
uxtb16 r6, r8 ; byte (two pixels) to halfwords
uxtb16 r7, r8, ror #8 ; another two pixels to halfwords
smlad r4, r6, r6, r4 ; dual signed multiply, add and accumulate (1)
; 2nd 4 pixels
ldr r6, [r2, #0x4] ; load 4 ref pixels
smlad r4, r7, r7, r4 ; dual signed multiply, add and accumulate (2)
usub8 r8, r5, r6 ; calculate difference
sel r7, r8, lr ; select bytes with positive difference
usub8 r9, r6, r5 ; calculate difference with reversed operands
sel r8, r9, lr ; select bytes with negative difference
; calculate partial sums
usad8 r5, r7, lr ; calculate sum of positive differences
usad8 r6, r8, lr ; calculate sum of negative differences
orr r8, r8, r7 ; differences of all 4 pixels
ldr r5, [r0, #0x8] ; load 4 src pixels
; calculate sse
uxtb16 r6, r8 ; byte (two pixels) to halfwords
uxtb16 r7, r8, ror #8 ; another two pixels to halfwords
smlad r4, r6, r6, r4 ; dual signed multiply, add and accumulate (1)
; 3rd 4 pixels
ldr r6, [r2, #0x8] ; load 4 ref pixels
smlad r4, r7, r7, r4 ; dual signed multiply, add and accumulate (2)
usub8 r8, r5, r6 ; calculate difference
sel r7, r8, lr ; select bytes with positive difference
usub8 r9, r6, r5 ; calculate difference with reversed operands
sel r8, r9, lr ; select bytes with negative difference
; calculate partial sums
usad8 r5, r7, lr ; calculate sum of positive differences
usad8 r6, r8, lr ; calculate sum of negative differences
orr r8, r8, r7 ; differences of all 4 pixels
ldr r5, [r0, #0xc] ; load 4 src pixels
; calculate sse
uxtb16 r6, r8 ; byte (two pixels) to halfwords
uxtb16 r7, r8, ror #8 ; another two pixels to halfwords
smlad r4, r6, r6, r4 ; dual signed multiply, add and accumulate (1)
; 4th 4 pixels
ldr r6, [r2, #0xc] ; load 4 ref pixels
smlad r4, r7, r7, r4 ; dual signed multiply, add and accumulate (2)
usub8 r8, r5, r6 ; calculate difference
add r0, r0, r1 ; set src_ptr to next row
sel r7, r8, lr ; select bytes with positive difference
usub8 r9, r6, r5 ; calculate difference with reversed operands
add r2, r2, r3 ; set dst_ptr to next row
sel r8, r9, lr ; select bytes with negative difference
; calculate partial sums
usad8 r5, r7, lr ; calculate sum of positive differences
usad8 r6, r8, lr ; calculate sum of negative differences
orr r8, r8, r7 ; differences of all 4 pixels
subs r12, r12, #1 ; next row
; calculate sse
uxtb16 r6, r8 ; byte (two pixels) to halfwords
uxtb16 r7, r8, ror #8 ; another two pixels to halfwords
smlad r4, r6, r6, r4 ; dual signed multiply, add and accumulate (1)
smlad r4, r7, r7, r4 ; dual signed multiply, add and accumulate (2)
bne loop
; return stuff
ldr r1, [sp, #28] ; get address of sse
mov r0, r4 ; return sse
str r4, [r1] ; store sse
pop {r4-r9, pc}
ENDP
END

131
media/libvpx/vp8/encoder/arm/neon/vp8_mse16x16_neon.c
Просмотреть файл

@ -1,131 +0,0 @@
/*
* Copyright (c) 2014 The WebM project authors. All Rights Reserved.
*
* Use of this source code is governed by a BSD-style license
* that can be found in the LICENSE file in the root of the source
* tree. An additional intellectual property rights grant can be found
* in the file PATENTS. All contributing project authors may
* be found in the AUTHORS file in the root of the source tree.
*/
#include <arm_neon.h>
unsigned int vp8_mse16x16_neon(
const unsigned char *src_ptr,
int source_stride,
const unsigned char *ref_ptr,
int recon_stride,
unsigned int *sse) {
int i;
int16x4_t d22s16, d23s16, d24s16, d25s16, d26s16, d27s16, d28s16, d29s16;
int64x1_t d0s64;
uint8x16_t q0u8, q1u8, q2u8, q3u8;
int32x4_t q7s32, q8s32, q9s32, q10s32;
uint16x8_t q11u16, q12u16, q13u16, q14u16;
int64x2_t q1s64;
q7s32 = vdupq_n_s32(0);
q8s32 = vdupq_n_s32(0);
q9s32 = vdupq_n_s32(0);
q10s32 = vdupq_n_s32(0);
for (i = 0; i < 8; i++) { // mse16x16_neon_loop
q0u8 = vld1q_u8(src_ptr);
src_ptr += source_stride;
q1u8 = vld1q_u8(src_ptr);
src_ptr += source_stride;
q2u8 = vld1q_u8(ref_ptr);
ref_ptr += recon_stride;
q3u8 = vld1q_u8(ref_ptr);
ref_ptr += recon_stride;
q11u16 = vsubl_u8(vget_low_u8(q0u8), vget_low_u8(q2u8));
q12u16 = vsubl_u8(vget_high_u8(q0u8), vget_high_u8(q2u8));
q13u16 = vsubl_u8(vget_low_u8(q1u8), vget_low_u8(q3u8));
q14u16 = vsubl_u8(vget_high_u8(q1u8), vget_high_u8(q3u8));
d22s16 = vreinterpret_s16_u16(vget_low_u16(q11u16));
d23s16 = vreinterpret_s16_u16(vget_high_u16(q11u16));
q7s32 = vmlal_s16(q7s32, d22s16, d22s16);
q8s32 = vmlal_s16(q8s32, d23s16, d23s16);
d24s16 = vreinterpret_s16_u16(vget_low_u16(q12u16));
d25s16 = vreinterpret_s16_u16(vget_high_u16(q12u16));
q9s32 = vmlal_s16(q9s32, d24s16, d24s16);
q10s32 = vmlal_s16(q10s32, d25s16, d25s16);
d26s16 = vreinterpret_s16_u16(vget_low_u16(q13u16));
d27s16 = vreinterpret_s16_u16(vget_high_u16(q13u16));
q7s32 = vmlal_s16(q7s32, d26s16, d26s16);
q8s32 = vmlal_s16(q8s32, d27s16, d27s16);
d28s16 = vreinterpret_s16_u16(vget_low_u16(q14u16));
d29s16 = vreinterpret_s16_u16(vget_high_u16(q14u16));
q9s32 = vmlal_s16(q9s32, d28s16, d28s16);
q10s32 = vmlal_s16(q10s32, d29s16, d29s16);
}
q7s32 = vaddq_s32(q7s32, q8s32);
q9s32 = vaddq_s32(q9s32, q10s32);
q10s32 = vaddq_s32(q7s32, q9s32);
q1s64 = vpaddlq_s32(q10s32);
d0s64 = vadd_s64(vget_low_s64(q1s64), vget_high_s64(q1s64));
vst1_lane_u32((uint32_t *)sse, vreinterpret_u32_s64(d0s64), 0);
return vget_lane_u32(vreinterpret_u32_s64(d0s64), 0);
}
unsigned int vp8_get4x4sse_cs_neon(
const unsigned char *src_ptr,
int source_stride,
const unsigned char *ref_ptr,
int recon_stride) {
int16x4_t d22s16, d24s16, d26s16, d28s16;
int64x1_t d0s64;
uint8x8_t d0u8, d1u8, d2u8, d3u8, d4u8, d5u8, d6u8, d7u8;
int32x4_t q7s32, q8s32, q9s32, q10s32;
uint16x8_t q11u16, q12u16, q13u16, q14u16;
int64x2_t q1s64;
d0u8 = vld1_u8(src_ptr);
src_ptr += source_stride;
d4u8 = vld1_u8(ref_ptr);
ref_ptr += recon_stride;
d1u8 = vld1_u8(src_ptr);
src_ptr += source_stride;
d5u8 = vld1_u8(ref_ptr);
ref_ptr += recon_stride;
d2u8 = vld1_u8(src_ptr);
src_ptr += source_stride;
d6u8 = vld1_u8(ref_ptr);
ref_ptr += recon_stride;
d3u8 = vld1_u8(src_ptr);
src_ptr += source_stride;
d7u8 = vld1_u8(ref_ptr);
ref_ptr += recon_stride;
q11u16 = vsubl_u8(d0u8, d4u8);
q12u16 = vsubl_u8(d1u8, d5u8);
q13u16 = vsubl_u8(d2u8, d6u8);
q14u16 = vsubl_u8(d3u8, d7u8);
d22s16 = vget_low_s16(vreinterpretq_s16_u16(q11u16));
d24s16 = vget_low_s16(vreinterpretq_s16_u16(q12u16));
d26s16 = vget_low_s16(vreinterpretq_s16_u16(q13u16));
d28s16 = vget_low_s16(vreinterpretq_s16_u16(q14u16));
q7s32 = vmull_s16(d22s16, d22s16);
q8s32 = vmull_s16(d24s16, d24s16);
q9s32 = vmull_s16(d26s16, d26s16);
q10s32 = vmull_s16(d28s16, d28s16);
q7s32 = vaddq_s32(q7s32, q8s32);
q9s32 = vaddq_s32(q9s32, q10s32);
q9s32 = vaddq_s32(q7s32, q9s32);
q1s64 = vpaddlq_s32(q9s32);
d0s64 = vadd_s64(vget_low_s64(q1s64), vget_high_s64(q1s64));
return vget_lane_u32(vreinterpret_u32_s64(d0s64), 0);
}

4
media/libvpx/vp8/encoder/bitstream.c
Просмотреть файл

@ -1543,7 +1543,7 @@ void vp8_pack_bitstream(VP8_COMP *cpi, unsigned char *dest, unsigned char * dest
if (pc->refresh_entropy_probs == 0)
{
/* save a copy for later refresh */
vpx_memcpy(&cpi->common.lfc, &cpi->common.fc, sizeof(cpi->common.fc));
memcpy(&cpi->common.lfc, &cpi->common.fc, sizeof(cpi->common.fc));
}
vp8_update_coef_probs(cpi);
@ -1620,7 +1620,7 @@ void vp8_pack_bitstream(VP8_COMP *cpi, unsigned char *dest, unsigned char * dest
/* concatenate partition buffers */
for(i = 0; i < num_part; i++)
{
vpx_memmove(dp, cpi->partition_d[i+1], cpi->partition_sz[i+1]);
memmove(dp, cpi->partition_d[i+1], cpi->partition_sz[i+1]);
cpi->partition_d[i+1] = dp;
dp += cpi->partition_sz[i+1];
}

2
media/libvpx/vp8/encoder/dct.c
Просмотреть файл

@ -11,6 +11,8 @@
#include <math.h>
#include "./vp8_rtcd.h"
void vp8_short_fdct4x4_c(short *input, short *output, int pitch)
{
int i;

14
media/libvpx/vp8/encoder/denoising.c
Просмотреть файл

@ -415,8 +415,8 @@ int vp8_denoiser_allocate(VP8_DENOISER *denoiser, int width, int height,
vp8_denoiser_free(denoiser);
return 1;
}
vpx_memset(denoiser->yv12_running_avg[i].buffer_alloc, 0,
denoiser->yv12_running_avg[i].frame_size);
memset(denoiser->yv12_running_avg[i].buffer_alloc, 0,
denoiser->yv12_running_avg[i].frame_size);
}
denoiser->yv12_mc_running_avg.flags = 0;
@ -428,19 +428,19 @@ int vp8_denoiser_allocate(VP8_DENOISER *denoiser, int width, int height,
return 1;
}
vpx_memset(denoiser->yv12_mc_running_avg.buffer_alloc, 0,
denoiser->yv12_mc_running_avg.frame_size);
memset(denoiser->yv12_mc_running_avg.buffer_alloc, 0,
denoiser->yv12_mc_running_avg.frame_size);
if (vp8_yv12_alloc_frame_buffer(&denoiser->yv12_last_source, width,
height, VP8BORDERINPIXELS) < 0) {
vp8_denoiser_free(denoiser);
return 1;
}
vpx_memset(denoiser->yv12_last_source.buffer_alloc, 0,
denoiser->yv12_last_source.frame_size);
memset(denoiser->yv12_last_source.buffer_alloc, 0,
denoiser->yv12_last_source.frame_size);
denoiser->denoise_state = vpx_calloc((num_mb_rows * num_mb_cols), 1);
vpx_memset(denoiser->denoise_state, 0, (num_mb_rows * num_mb_cols));
memset(denoiser->denoise_state, 0, (num_mb_rows * num_mb_cols));
vp8_denoiser_set_parameters(denoiser, mode);
denoiser->nmse_source_diff = 0;
denoiser->nmse_source_diff_count = 0;

14
media/libvpx/vp8/encoder/encodeframe.c
Просмотреть файл

@ -11,6 +11,7 @@
#include "vpx_config.h"
#include "vp8_rtcd.h"
#include "./vpx_dsp_rtcd.h"
#include "encodemb.h"
#include "encodemv.h"
#include "vp8/common/common.h"
@ -90,7 +91,7 @@ static unsigned int tt_activity_measure( VP8_COMP *cpi, MACROBLOCK *x )
* lambda using a non-linear combination (e.g., the smallest, or second
* smallest, etc.).
*/
act = vp8_variance16x16(x->src.y_buffer,
act = vpx_variance16x16(x->src.y_buffer,
x->src.y_stride, VP8_VAR_OFFS, 0, &sse);
act = act<<4;
@ -155,8 +156,8 @@ static void calc_av_activity( VP8_COMP *cpi, int64_t activity_sum )
cpi->common.MBs));
/* Copy map to sort list */
vpx_memcpy( sortlist, cpi->mb_activity_map,
sizeof(unsigned int) * cpi->common.MBs );
memcpy( sortlist, cpi->mb_activity_map,
sizeof(unsigned int) * cpi->common.MBs );
/* Ripple each value down to its correct position */
@ -665,8 +666,7 @@ static void init_encode_frame_mb_context(VP8_COMP *cpi)
x->mvc = cm->fc.mvc;
vpx_memset(cm->above_context, 0,
sizeof(ENTROPY_CONTEXT_PLANES) * cm->mb_cols);
memset(cm->above_context, 0, sizeof(ENTROPY_CONTEXT_PLANES) * cm->mb_cols);
/* Special case treatment when GF and ARF are not sensible options
* for reference
@ -744,7 +744,7 @@ void vp8_encode_frame(VP8_COMP *cpi)
const int num_part = (1 << cm->multi_token_partition);
#endif
vpx_memset(segment_counts, 0, sizeof(segment_counts));
memset(segment_counts, 0, sizeof(segment_counts));
totalrate = 0;
if (cpi->compressor_speed == 2)
@ -974,7 +974,7 @@ void vp8_encode_frame(VP8_COMP *cpi)
int i;
/* Set to defaults */
vpx_memset(xd->mb_segment_tree_probs, 255 , sizeof(xd->mb_segment_tree_probs));
memset(xd->mb_segment_tree_probs, 255 , sizeof(xd->mb_segment_tree_probs));
tot_count = segment_counts[0] + segment_counts[1] + segment_counts[2] + segment_counts[3];

3
media/libvpx/vp8/encoder/encodeintra.c
Просмотреть файл

@ -11,6 +11,7 @@
#include "vpx_config.h"
#include "vp8_rtcd.h"
#include "./vpx_dsp_rtcd.h"
#include "quantize.h"
#include "vp8/common/reconintra4x4.h"
#include "encodemb.h"
@ -44,7 +45,7 @@ int vp8_encode_intra(VP8_COMP *cpi, MACROBLOCK *x, int use_dc_pred)
}
}
intra_pred_var = vp8_get_mb_ss(x->src_diff);
intra_pred_var = vpx_get_mb_ss(x->src_diff);
return intra_pred_var;
}

12
media/libvpx/vp8/encoder/encodemb.c
Просмотреть файл

@ -506,8 +506,8 @@ static void optimize_mb(MACROBLOCK *x)
ENTROPY_CONTEXT *ta;
ENTROPY_CONTEXT *tl;
vpx_memcpy(&t_above, x->e_mbd.above_context, sizeof(ENTROPY_CONTEXT_PLANES));
vpx_memcpy(&t_left, x->e_mbd.left_context, sizeof(ENTROPY_CONTEXT_PLANES));
memcpy(&t_above, x->e_mbd.above_context, sizeof(ENTROPY_CONTEXT_PLANES));
memcpy(&t_left, x->e_mbd.left_context, sizeof(ENTROPY_CONTEXT_PLANES));
ta = (ENTROPY_CONTEXT *)&t_above;
tl = (ENTROPY_CONTEXT *)&t_left;
@ -555,8 +555,8 @@ void vp8_optimize_mby(MACROBLOCK *x)
if (!x->e_mbd.left_context)
return;
vpx_memcpy(&t_above, x->e_mbd.above_context, sizeof(ENTROPY_CONTEXT_PLANES));
vpx_memcpy(&t_left, x->e_mbd.left_context, sizeof(ENTROPY_CONTEXT_PLANES));
memcpy(&t_above, x->e_mbd.above_context, sizeof(ENTROPY_CONTEXT_PLANES));
memcpy(&t_left, x->e_mbd.left_context, sizeof(ENTROPY_CONTEXT_PLANES));
ta = (ENTROPY_CONTEXT *)&t_above;
tl = (ENTROPY_CONTEXT *)&t_left;
@ -595,8 +595,8 @@ void vp8_optimize_mbuv(MACROBLOCK *x)
if (!x->e_mbd.left_context)
return;
vpx_memcpy(&t_above, x->e_mbd.above_context, sizeof(ENTROPY_CONTEXT_PLANES));
vpx_memcpy(&t_left, x->e_mbd.left_context, sizeof(ENTROPY_CONTEXT_PLANES));
memcpy(&t_above, x->e_mbd.above_context, sizeof(ENTROPY_CONTEXT_PLANES));
memcpy(&t_left, x->e_mbd.left_context, sizeof(ENTROPY_CONTEXT_PLANES));
ta = (ENTROPY_CONTEXT *)&t_above;
tl = (ENTROPY_CONTEXT *)&t_left;

26
media/libvpx/vp8/encoder/ethreading.c
Просмотреть файл

@ -19,8 +19,6 @@
extern void vp8cx_mb_init_quantizer(VP8_COMP *cpi, MACROBLOCK *x, int ok_to_skip);
extern void vp8_loopfilter_frame(VP8_COMP *cpi, VP8_COMMON *cm);
static THREAD_FUNCTION thread_loopfilter(void *p_data)
{
VP8_COMP *cpi = (VP8_COMP *)(((LPFTHREAD_DATA *)p_data)->ptr1);
@ -416,14 +414,13 @@ static void setup_mbby_copy(MACROBLOCK *mbdst, MACROBLOCK *mbsrc)
zd->subpixel_predict16x16 = xd->subpixel_predict16x16;
zd->segmentation_enabled = xd->segmentation_enabled;
zd->mb_segement_abs_delta = xd->mb_segement_abs_delta;
vpx_memcpy(zd->segment_feature_data, xd->segment_feature_data,
sizeof(xd->segment_feature_data));
memcpy(zd->segment_feature_data, xd->segment_feature_data,
sizeof(xd->segment_feature_data));
vpx_memcpy(zd->dequant_y1_dc, xd->dequant_y1_dc,
sizeof(xd->dequant_y1_dc));
vpx_memcpy(zd->dequant_y1, xd->dequant_y1, sizeof(xd->dequant_y1));
vpx_memcpy(zd->dequant_y2, xd->dequant_y2, sizeof(xd->dequant_y2));
vpx_memcpy(zd->dequant_uv, xd->dequant_uv, sizeof(xd->dequant_uv));
memcpy(zd->dequant_y1_dc, xd->dequant_y1_dc, sizeof(xd->dequant_y1_dc));
memcpy(zd->dequant_y1, xd->dequant_y1, sizeof(xd->dequant_y1));
memcpy(zd->dequant_y2, xd->dequant_y2, sizeof(xd->dequant_y2));
memcpy(zd->dequant_uv, xd->dequant_uv, sizeof(xd->dequant_uv));
#if 1
/*TODO: Remove dequant from BLOCKD. This is a temporary solution until
@ -438,15 +435,14 @@ static void setup_mbby_copy(MACROBLOCK *mbdst, MACROBLOCK *mbsrc)
#endif
vpx_memcpy(z->rd_threshes, x->rd_threshes, sizeof(x->rd_threshes));
vpx_memcpy(z->rd_thresh_mult, x->rd_thresh_mult,
sizeof(x->rd_thresh_mult));
memcpy(z->rd_threshes, x->rd_threshes, sizeof(x->rd_threshes));
memcpy(z->rd_thresh_mult, x->rd_thresh_mult, sizeof(x->rd_thresh_mult));
z->zbin_over_quant = x->zbin_over_quant;
z->zbin_mode_boost_enabled = x->zbin_mode_boost_enabled;
z->zbin_mode_boost = x->zbin_mode_boost;
vpx_memset(z->error_bins, 0, sizeof(z->error_bins));
memset(z->error_bins, 0, sizeof(z->error_bins));
}
}
@ -472,7 +468,7 @@ void vp8cx_init_mbrthread_data(VP8_COMP *cpi,
mbd->subpixel_predict16x16 = xd->subpixel_predict16x16;
mb->gf_active_ptr = x->gf_active_ptr;
vpx_memset(mbr_ei[i].segment_counts, 0, sizeof(mbr_ei[i].segment_counts));
memset(mbr_ei[i].segment_counts, 0, sizeof(mbr_ei[i].segment_counts));
mbr_ei[i].totalrate = 0;
mb->partition_info = x->pi + x->e_mbd.mode_info_stride * (i + 1);
@ -547,7 +543,7 @@ int vp8cx_create_encoder_threads(VP8_COMP *cpi)
vpx_malloc(sizeof(sem_t) * th_count));
CHECK_MEM_ERROR(cpi->mb_row_ei,
vpx_memalign(32, sizeof(MB_ROW_COMP) * th_count));
vpx_memset(cpi->mb_row_ei, 0, sizeof(MB_ROW_COMP) * th_count);
memset(cpi->mb_row_ei, 0, sizeof(MB_ROW_COMP) * th_count);
CHECK_MEM_ERROR(cpi->en_thread_data,
vpx_malloc(sizeof(ENCODETHREAD_DATA) * th_count));

44
media/libvpx/vp8/encoder/firstpass.c
Просмотреть файл

@ -12,6 +12,7 @@
#include <limits.h>
#include <stdio.h>
#include "./vpx_dsp_rtcd.h"
#include "./vpx_scale_rtcd.h"
#include "block.h"
#include "onyx_int.h"
@ -34,8 +35,6 @@
/* #define OUTPUT_FPF 1 */
extern void vp8cx_frame_init_quantizer(VP8_COMP *cpi);
extern void vp8_set_mbmode_and_mvs(MACROBLOCK *x, MB_PREDICTION_MODE mb, int_mv *mv);
extern void vp8_alloc_compressor_data(VP8_COMP *cpi);
#define GFQ_ADJUSTMENT vp8_gf_boost_qadjustment[Q]
extern int vp8_kf_boost_qadjustment[QINDEX_RANGE];
@ -424,14 +423,14 @@ static void zz_motion_search( VP8_COMP *cpi, MACROBLOCK * x,
/* Set up pointers for this macro block raw buffer */
raw_ptr = (unsigned char *)(raw_buffer->y_buffer + recon_yoffset
+ d->offset);
vp8_mse16x16 ( src_ptr, src_stride, raw_ptr, raw_stride,
(unsigned int *)(raw_motion_err));
vpx_mse16x16(src_ptr, src_stride, raw_ptr, raw_stride,
(unsigned int *)(raw_motion_err));
/* Set up pointers for this macro block recon buffer */
xd->pre.y_buffer = recon_buffer->y_buffer + recon_yoffset;
ref_ptr = (unsigned char *)(xd->pre.y_buffer + d->offset );
vp8_mse16x16 ( src_ptr, src_stride, ref_ptr, ref_stride,
(unsigned int *)(best_motion_err));
vpx_mse16x16(src_ptr, src_stride, ref_ptr, ref_stride,
(unsigned int *)(best_motion_err));
}
static void first_pass_motion_search(VP8_COMP *cpi, MACROBLOCK *x,
@ -455,7 +454,7 @@ static void first_pass_motion_search(VP8_COMP *cpi, MACROBLOCK *x,
int new_mv_mode_penalty = 256;
/* override the default variance function to use MSE */
v_fn_ptr.vf = vp8_mse16x16;
v_fn_ptr.vf = vpx_mse16x16;
/* Set up pointers for this macro block recon buffer */
xd->pre.y_buffer = recon_buffer->y_buffer + recon_yoffset;
@ -573,7 +572,7 @@ void vp8_first_pass(VP8_COMP *cpi)
{
int flag[2] = {1, 1};
vp8_initialize_rd_consts(cpi, x, vp8_dc_quant(cm->base_qindex, cm->y1dc_delta_q));
vpx_memcpy(cm->fc.mvc, vp8_default_mv_context, sizeof(vp8_default_mv_context));
memcpy(cm->fc.mvc, vp8_default_mv_context, sizeof(vp8_default_mv_context));
vp8_build_component_cost_table(cpi->mb.mvcost, (const MV_CONTEXT *) cm->fc.mvc, flag);
}
@ -1329,8 +1328,6 @@ static int estimate_kf_group_q(VP8_COMP *cpi, double section_err, int section_ta
return Q;
}
extern void vp8_new_framerate(VP8_COMP *cpi, double framerate);
void vp8_init_second_pass(VP8_COMP *cpi)
{
FIRSTPASS_STATS this_frame;
@ -1779,7 +1776,7 @@ static void define_gf_group(VP8_COMP *cpi, FIRSTPASS_STATS *this_frame)
start_pos = cpi->twopass.stats_in;
vpx_memset(&next_frame, 0, sizeof(next_frame)); /* assure clean */
memset(&next_frame, 0, sizeof(next_frame)); /* assure clean */
/* Load stats for the current frame. */
mod_frame_err = calculate_modified_err(cpi, this_frame);
@ -1875,7 +1872,7 @@ static void define_gf_group(VP8_COMP *cpi, FIRSTPASS_STATS *this_frame)
break;
}
vpx_memcpy(this_frame, &next_frame, sizeof(*this_frame));
memcpy(this_frame, &next_frame, sizeof(*this_frame));
old_boost_score = boost_score;
}
@ -2445,7 +2442,7 @@ void vp8_second_pass(VP8_COMP *cpi)
if (cpi->twopass.frames_to_key == 0)
{
/* Define next KF group and assign bits to it */
vpx_memcpy(&this_frame_copy, &this_frame, sizeof(this_frame));
memcpy(&this_frame_copy, &this_frame, sizeof(this_frame));
find_next_key_frame(cpi, &this_frame_copy);
/* Special case: Error error_resilient_mode mode does not make much
@ -2471,7 +2468,7 @@ void vp8_second_pass(VP8_COMP *cpi)
if (cpi->frames_till_gf_update_due == 0)
{
/* Define next gf group and assign bits to it */
vpx_memcpy(&this_frame_copy, &this_frame, sizeof(this_frame));
memcpy(&this_frame_copy, &this_frame, sizeof(this_frame));
define_gf_group(cpi, &this_frame_copy);
/* If we are going to code an altref frame at the end of the group
@ -2487,7 +2484,7 @@ void vp8_second_pass(VP8_COMP *cpi)
* to the GF group
*/
int bak = cpi->per_frame_bandwidth;
vpx_memcpy(&this_frame_copy, &this_frame, sizeof(this_frame));
memcpy(&this_frame_copy, &this_frame, sizeof(this_frame));
assign_std_frame_bits(cpi, &this_frame_copy);
cpi->per_frame_bandwidth = bak;
}
@ -2510,14 +2507,14 @@ void vp8_second_pass(VP8_COMP *cpi)
if (cpi->common.frame_type != KEY_FRAME)
{
/* Assign bits from those allocated to the GF group */
vpx_memcpy(&this_frame_copy, &this_frame, sizeof(this_frame));
memcpy(&this_frame_copy, &this_frame, sizeof(this_frame));
assign_std_frame_bits(cpi, &this_frame_copy);
}
}
else
{
/* Assign bits from those allocated to the GF group */
vpx_memcpy(&this_frame_copy, &this_frame, sizeof(this_frame));
memcpy(&this_frame_copy, &this_frame, sizeof(this_frame));
assign_std_frame_bits(cpi, &this_frame_copy);
}
}
@ -2658,7 +2655,7 @@ static int test_candidate_kf(VP8_COMP *cpi, FIRSTPASS_STATS *last_frame, FIRSTP
double decay_accumulator = 1.0;
double next_iiratio;
vpx_memcpy(&local_next_frame, next_frame, sizeof(*next_frame));
memcpy(&local_next_frame, next_frame, sizeof(*next_frame));
/* Note the starting file position so we can reset to it */
start_pos = cpi->twopass.stats_in;
@ -2735,7 +2732,7 @@ static void find_next_key_frame(VP8_COMP *cpi, FIRSTPASS_STATS *this_frame)
double kf_group_coded_err = 0.0;
double recent_loop_decay[8] = {1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0};
vpx_memset(&next_frame, 0, sizeof(next_frame));
memset(&next_frame, 0, sizeof(next_frame));
vp8_clear_system_state();
start_position = cpi->twopass.stats_in;
@ -2756,7 +2753,7 @@ static void find_next_key_frame(VP8_COMP *cpi, FIRSTPASS_STATS *this_frame)
cpi->twopass.frames_to_key = 1;
/* Take a copy of the initial frame details */
vpx_memcpy(&first_frame, this_frame, sizeof(*this_frame));
memcpy(&first_frame, this_frame, sizeof(*this_frame));
cpi->twopass.kf_group_bits = 0;
cpi->twopass.kf_group_error_left = 0;
@ -2779,7 +2776,7 @@ static void find_next_key_frame(VP8_COMP *cpi, FIRSTPASS_STATS *this_frame)
kf_group_coded_err += this_frame->coded_error;
/* Load the next frame's stats. */
vpx_memcpy(&last_frame, this_frame, sizeof(*this_frame));
memcpy(&last_frame, this_frame, sizeof(*this_frame));
input_stats(cpi, this_frame);
/* Provided that we are not at the end of the file... */
@ -2847,7 +2844,7 @@ static void find_next_key_frame(VP8_COMP *cpi, FIRSTPASS_STATS *this_frame)
cpi->twopass.frames_to_key /= 2;
/* Copy first frame details */
vpx_memcpy(&tmp_frame, &first_frame, sizeof(first_frame));
memcpy(&tmp_frame, &first_frame, sizeof(first_frame));
/* Reset to the start of the group */
reset_fpf_position(cpi, start_position);
@ -2969,7 +2966,6 @@ static void find_next_key_frame(VP8_COMP *cpi, FIRSTPASS_STATS *this_frame)
*/
decay_accumulator = 1.0;
boost_score = 0.0;
loop_decay_rate = 1.00; /* Starting decay rate */
for (i = 0 ; i < cpi->twopass.frames_to_key ; i++)
{
@ -3213,7 +3209,7 @@ static void find_next_key_frame(VP8_COMP *cpi, FIRSTPASS_STATS *this_frame)
int new_width = cpi->oxcf.Width;
int new_height = cpi->oxcf.Height;
int projected_buffer_level = (int)cpi->buffer_level;
int projected_buffer_level;
int tmp_q;
double projected_bits_perframe;

56
media/libvpx/vp8/encoder/mcomp.c
Просмотреть файл

@ -9,6 +9,8 @@
*/
#include "./vp8_rtcd.h"
#include "./vpx_dsp_rtcd.h"
#include "onyx_int.h"
#include "mcomp.h"
#include "vpx_mem/vpx_mem.h"
@ -900,7 +902,7 @@ int vp8_hex_search
this_offset = base_offset + (br * (pre_stride)) + bc;
this_mv.as_mv.row = br;
this_mv.as_mv.col = bc;
bestsad = vfp->sdf(what, what_stride, this_offset, in_what_stride, UINT_MAX)
bestsad = vfp->sdf(what, what_stride, this_offset, in_what_stride)
+ mvsad_err_cost(&this_mv, &fcenter_mv, mvsadcost, sad_per_bit);
#if CONFIG_MULTI_RES_ENCODING
@ -927,7 +929,7 @@ int vp8_hex_search
this_mv.as_mv.row = br + hex[i].row;
this_mv.as_mv.col = bc + hex[i].col;
this_offset = base_offset + (this_mv.as_mv.row * in_what_stride) + this_mv.as_mv.col;
thissad = vfp->sdf(what, what_stride, this_offset, in_what_stride, bestsad);
thissad = vfp->sdf(what, what_stride, this_offset, in_what_stride);
CHECK_BETTER
}
}else
@ -938,7 +940,7 @@ int vp8_hex_search
this_mv.as_mv.col = bc + hex[i].col;
CHECK_POINT
this_offset = base_offset + (this_mv.as_mv.row * in_what_stride) + this_mv.as_mv.col;
thissad = vfp->sdf(what, what_stride, this_offset, in_what_stride, bestsad);
thissad = vfp->sdf(what, what_stride, this_offset, in_what_stride);
CHECK_BETTER
}
}
@ -964,7 +966,7 @@ int vp8_hex_search
this_mv.as_mv.row = br + next_chkpts[k][i].row;
this_mv.as_mv.col = bc + next_chkpts[k][i].col;
this_offset = base_offset + (this_mv.as_mv.row * (in_what_stride)) + this_mv.as_mv.col;
thissad = vfp->sdf(what, what_stride, this_offset, in_what_stride, bestsad);
thissad = vfp->sdf(what, what_stride, this_offset, in_what_stride);
CHECK_BETTER
}
}else
@ -975,7 +977,7 @@ int vp8_hex_search
this_mv.as_mv.col = bc + next_chkpts[k][i].col;
CHECK_POINT
this_offset = base_offset + (this_mv.as_mv.row * (in_what_stride)) + this_mv.as_mv.col;
thissad = vfp->sdf(what, what_stride, this_offset, in_what_stride, bestsad);
thissad = vfp->sdf(what, what_stride, this_offset, in_what_stride);
CHECK_BETTER
}
}
@ -1006,7 +1008,7 @@ cal_neighbors:
this_mv.as_mv.row = br + neighbors[i].row;
this_mv.as_mv.col = bc + neighbors[i].col;
this_offset = base_offset + (this_mv.as_mv.row * (in_what_stride)) + this_mv.as_mv.col;
thissad = vfp->sdf(what, what_stride, this_offset, in_what_stride, bestsad);
thissad = vfp->sdf(what, what_stride, this_offset, in_what_stride);
CHECK_BETTER
}
}else
@ -1017,7 +1019,7 @@ cal_neighbors:
this_mv.as_mv.col = bc + neighbors[i].col;
CHECK_POINT
this_offset = base_offset + (this_mv.as_mv.row * (in_what_stride)) + this_mv.as_mv.col;
thissad = vfp->sdf(what, what_stride, this_offset, in_what_stride, bestsad);
thissad = vfp->sdf(what, what_stride, this_offset, in_what_stride);
CHECK_BETTER
}
}
@ -1101,7 +1103,7 @@ int vp8_diamond_search_sad_c
best_address = in_what;
/* Check the starting position */
bestsad = fn_ptr->sdf(what, what_stride, in_what, in_what_stride, UINT_MAX)
bestsad = fn_ptr->sdf(what, what_stride, in_what, in_what_stride)
+ mvsad_err_cost(best_mv, &fcenter_mv, mvsadcost, sad_per_bit);
/* search_param determines the length of the initial step and hence
@ -1126,7 +1128,7 @@ int vp8_diamond_search_sad_c
{
check_here = ss[i].offset + best_address;
thissad = fn_ptr->sdf(what, what_stride, check_here, in_what_stride, bestsad);
thissad = fn_ptr->sdf(what, what_stride, check_here, in_what_stride);
if (thissad < bestsad)
{
@ -1225,7 +1227,7 @@ int vp8_diamond_search_sadx4
best_address = in_what;
/* Check the starting position */
bestsad = fn_ptr->sdf(what, what_stride, in_what, in_what_stride, UINT_MAX)
bestsad = fn_ptr->sdf(what, what_stride, in_what, in_what_stride)
+ mvsad_err_cost(best_mv, &fcenter_mv, mvsadcost, sad_per_bit);
/* search_param determines the length of the initial step and hence the
@ -1293,7 +1295,7 @@ int vp8_diamond_search_sadx4
(this_row_offset > x->mv_row_min) && (this_row_offset < x->mv_row_max))
{
check_here = ss[i].offset + best_address;
thissad = fn_ptr->sdf(what, what_stride, check_here, in_what_stride, bestsad);
thissad = fn_ptr->sdf(what, what_stride, check_here, in_what_stride);
if (thissad < bestsad)
{
@ -1376,8 +1378,7 @@ int vp8_full_search_sad_c(MACROBLOCK *x, BLOCK *b, BLOCKD *d, int_mv *ref_mv,
best_mv->as_mv.col = ref_col;
/* Baseline value at the centre */
bestsad = fn_ptr->sdf(what, what_stride, bestaddress,
in_what_stride, UINT_MAX)
bestsad = fn_ptr->sdf(what, what_stride, bestaddress, in_what_stride)
+ mvsad_err_cost(best_mv, &fcenter_mv, mvsadcost, sad_per_bit);
/* Apply further limits to prevent us looking using vectors that
@ -1402,7 +1403,7 @@ int vp8_full_search_sad_c(MACROBLOCK *x, BLOCK *b, BLOCKD *d, int_mv *ref_mv,
for (c = col_min; c < col_max; c++)
{
thissad = fn_ptr->sdf(what, what_stride, check_here, in_what_stride, bestsad);
thissad = fn_ptr->sdf(what, what_stride, check_here, in_what_stride);
this_mv.as_mv.col = c;
thissad += mvsad_err_cost(&this_mv, &fcenter_mv,
@ -1474,8 +1475,7 @@ int vp8_full_search_sadx3(MACROBLOCK *x, BLOCK *b, BLOCKD *d, int_mv *ref_mv,
best_mv->as_mv.col = ref_col;
/* Baseline value at the centre */
bestsad = fn_ptr->sdf(what, what_stride, bestaddress,
in_what_stride, UINT_MAX)
bestsad = fn_ptr->sdf(what, what_stride, bestaddress, in_what_stride)
+ mvsad_err_cost(best_mv, &fcenter_mv, mvsadcost, sad_per_bit);
/* Apply further limits to prevent us looking using vectors that stretch
@ -1531,7 +1531,7 @@ int vp8_full_search_sadx3(MACROBLOCK *x, BLOCK *b, BLOCKD *d, int_mv *ref_mv,
while (c < col_max)
{
thissad = fn_ptr->sdf(what, what_stride, check_here, in_what_stride, bestsad);
thissad = fn_ptr->sdf(what, what_stride, check_here, in_what_stride);
if (thissad < bestsad)
{
@ -1590,7 +1590,8 @@ int vp8_full_search_sadx8(MACROBLOCK *x, BLOCK *b, BLOCKD *d, int_mv *ref_mv,
int col_min = ref_col - distance;
int col_max = ref_col + distance;
DECLARE_ALIGNED_ARRAY(16, unsigned short, sad_array8, 8);
// TODO(johannkoenig): check if this alignment is necessary.
DECLARE_ALIGNED(16, unsigned int, sad_array8[8]);
unsigned int sad_array[3];
int *mvsadcost[2];
@ -1609,8 +1610,7 @@ int vp8_full_search_sadx8(MACROBLOCK *x, BLOCK *b, BLOCKD *d, int_mv *ref_mv,
best_mv->as_mv.col = ref_col;
/* Baseline value at the centre */
bestsad = fn_ptr->sdf(what, what_stride,
bestaddress, in_what_stride, UINT_MAX)
bestsad = fn_ptr->sdf(what, what_stride, bestaddress, in_what_stride)
+ mvsad_err_cost(best_mv, &fcenter_mv, mvsadcost, sad_per_bit);
/* Apply further limits to prevent us looking using vectors that stretch
@ -1696,7 +1696,7 @@ int vp8_full_search_sadx8(MACROBLOCK *x, BLOCK *b, BLOCKD *d, int_mv *ref_mv,
while (c < col_max)
{
thissad = fn_ptr->sdf(what, what_stride, check_here , in_what_stride, bestsad);
thissad = fn_ptr->sdf(what, what_stride, check_here , in_what_stride);
if (thissad < bestsad)
{
@ -1754,8 +1754,7 @@ int vp8_refining_search_sad_c(MACROBLOCK *x, BLOCK *b, BLOCKD *d, int_mv *ref_mv
fcenter_mv.as_mv.row = center_mv->as_mv.row >> 3;
fcenter_mv.as_mv.col = center_mv->as_mv.col >> 3;
bestsad = fn_ptr->sdf(what, what_stride, best_address,
in_what_stride, UINT_MAX)
bestsad = fn_ptr->sdf(what, what_stride, best_address, in_what_stride)
+ mvsad_err_cost(ref_mv, &fcenter_mv, mvsadcost, error_per_bit);
for (i=0; i<search_range; i++)
@ -1771,7 +1770,7 @@ int vp8_refining_search_sad_c(MACROBLOCK *x, BLOCK *b, BLOCKD *d, int_mv *ref_mv
(this_row_offset > x->mv_row_min) && (this_row_offset < x->mv_row_max))
{
check_here = (neighbors[j].row)*in_what_stride + neighbors[j].col + best_address;
thissad = fn_ptr->sdf(what, what_stride, check_here , in_what_stride, bestsad);
thissad = fn_ptr->sdf(what, what_stride, check_here , in_what_stride);
if (thissad < bestsad)
{
@ -1834,8 +1833,7 @@ int vp8_refining_search_sadx4(MACROBLOCK *x, BLOCK *b, BLOCKD *d,
fcenter_mv.as_mv.row = center_mv->as_mv.row >> 3;
fcenter_mv.as_mv.col = center_mv->as_mv.col >> 3;
bestsad = fn_ptr->sdf(what, what_stride, best_address,
in_what_stride, UINT_MAX)
bestsad = fn_ptr->sdf(what, what_stride, best_address, in_what_stride)
+ mvsad_err_cost(ref_mv, &fcenter_mv, mvsadcost, error_per_bit);
for (i=0; i<search_range; i++)
@ -1886,7 +1884,7 @@ int vp8_refining_search_sadx4(MACROBLOCK *x, BLOCK *b, BLOCKD *d,
(this_row_offset > x->mv_row_min) && (this_row_offset < x->mv_row_max))
{
check_here = (neighbors[j].row)*in_what_stride + neighbors[j].col + best_address;
thissad = fn_ptr->sdf(what, what_stride, check_here , in_what_stride, bestsad);
thissad = fn_ptr->sdf(what, what_stride, check_here , in_what_stride);
if (thissad < bestsad)
{
@ -1978,8 +1976,8 @@ void print_mode_context(void)
#ifdef VP8_ENTROPY_STATS
void init_mv_ref_counts()
{
vpx_memset(mv_ref_ct, 0, sizeof(mv_ref_ct));
vpx_memset(mv_mode_cts, 0, sizeof(mv_mode_cts));
memset(mv_ref_ct, 0, sizeof(mv_ref_ct));
memset(mv_mode_cts, 0, sizeof(mv_mode_cts));
}
void accum_mv_refs(MB_PREDICTION_MODE m, const int ct[4])

1
media/libvpx/vp8/encoder/modecosts.c
Просмотреть файл

@ -10,6 +10,7 @@
#include "vp8/common/blockd.h"
#include "modecosts.h"
#include "onyx_int.h"
#include "treewriter.h"
#include "vp8/common/entropymode.h"

4
media/libvpx/vp8/encoder/modecosts.h
Просмотреть файл

@ -16,7 +16,9 @@
extern "C" {
#endif
void vp8_init_mode_costs(VP8_COMP *x);
struct VP8_COMP;
void vp8_init_mode_costs(struct VP8_COMP *x);
#ifdef __cplusplus
} // extern "C"

129
media/libvpx/vp8/encoder/onyx_if.c
Просмотреть файл

@ -11,6 +11,7 @@
#include "vpx_config.h"
#include "./vpx_scale_rtcd.h"
#include "./vpx_dsp_rtcd.h"
#include "./vp8_rtcd.h"
#include "vp8/common/onyxc_int.h"
#include "vp8/common/blockd.h"
@ -428,10 +429,10 @@ static void setup_features(VP8_COMP *cpi)
cpi->mb.e_mbd.mode_ref_lf_delta_enabled = 0;
cpi->mb.e_mbd.mode_ref_lf_delta_update = 0;
vpx_memset(cpi->mb.e_mbd.ref_lf_deltas, 0, sizeof(cpi->mb.e_mbd.ref_lf_deltas));
vpx_memset(cpi->mb.e_mbd.mode_lf_deltas, 0, sizeof(cpi->mb.e_mbd.mode_lf_deltas));
vpx_memset(cpi->mb.e_mbd.last_ref_lf_deltas, 0, sizeof(cpi->mb.e_mbd.ref_lf_deltas));
vpx_memset(cpi->mb.e_mbd.last_mode_lf_deltas, 0, sizeof(cpi->mb.e_mbd.mode_lf_deltas));
memset(cpi->mb.e_mbd.ref_lf_deltas, 0, sizeof(cpi->mb.e_mbd.ref_lf_deltas));
memset(cpi->mb.e_mbd.mode_lf_deltas, 0, sizeof(cpi->mb.e_mbd.mode_lf_deltas));
memset(cpi->mb.e_mbd.last_ref_lf_deltas, 0, sizeof(cpi->mb.e_mbd.ref_lf_deltas));
memset(cpi->mb.e_mbd.last_mode_lf_deltas, 0, sizeof(cpi->mb.e_mbd.mode_lf_deltas));
set_default_lf_deltas(cpi);
@ -508,7 +509,7 @@ static void disable_segmentation(VP8_COMP *cpi)
static void set_segmentation_map(VP8_COMP *cpi, unsigned char *segmentation_map)
{
/* Copy in the new segmentation map */
vpx_memcpy(cpi->segmentation_map, segmentation_map, (cpi->common.mb_rows * cpi->common.mb_cols));
memcpy(cpi->segmentation_map, segmentation_map, (cpi->common.mb_rows * cpi->common.mb_cols));
/* Signal that the map should be updated. */
cpi->mb.e_mbd.update_mb_segmentation_map = 1;
@ -530,7 +531,7 @@ static void set_segmentation_map(VP8_COMP *cpi, unsigned char *segmentation_map)
static void set_segment_data(VP8_COMP *cpi, signed char *feature_data, unsigned char abs_delta)
{
cpi->mb.e_mbd.mb_segement_abs_delta = abs_delta;
vpx_memcpy(cpi->segment_feature_data, feature_data, sizeof(cpi->segment_feature_data));
memcpy(cpi->segment_feature_data, feature_data, sizeof(cpi->segment_feature_data));
}
@ -586,7 +587,8 @@ static void cyclic_background_refresh(VP8_COMP *cpi, int Q, int lf_adjustment)
// Turn-off under certain conditions (i.e., away from key frame, and if
// we are at good quality (low Q) and most of the blocks were skipped-encoded
// in previous frame.
if (Q >= 100) {
int qp_thresh = (cpi->oxcf.screen_content_mode == 2) ? 80 : 100;
if (Q >= qp_thresh) {
cpi->cyclic_refresh_mode_max_mbs_perframe =
(cpi->common.mb_rows * cpi->common.mb_cols) / 10;
} else if (cpi->frames_since_key > 250 &&
@ -602,7 +604,7 @@ static void cyclic_background_refresh(VP8_COMP *cpi, int Q, int lf_adjustment)
// Set every macroblock to be eligible for update.
// For key frame this will reset seg map to 0.
vpx_memset(cpi->segmentation_map, 0, mbs_in_frame);
memset(cpi->segmentation_map, 0, mbs_in_frame);
if (cpi->common.frame_type != KEY_FRAME && block_count > 0)
{
@ -686,8 +688,8 @@ static void set_default_lf_deltas(VP8_COMP *cpi)
cpi->mb.e_mbd.mode_ref_lf_delta_enabled = 1;
cpi->mb.e_mbd.mode_ref_lf_delta_update = 1;
vpx_memset(cpi->mb.e_mbd.ref_lf_deltas, 0, sizeof(cpi->mb.e_mbd.ref_lf_deltas));
vpx_memset(cpi->mb.e_mbd.mode_lf_deltas, 0, sizeof(cpi->mb.e_mbd.mode_lf_deltas));
memset(cpi->mb.e_mbd.ref_lf_deltas, 0, sizeof(cpi->mb.e_mbd.ref_lf_deltas));
memset(cpi->mb.e_mbd.mode_lf_deltas, 0, sizeof(cpi->mb.e_mbd.mode_lf_deltas));
/* Test of ref frame deltas */
cpi->mb.e_mbd.ref_lf_deltas[INTRA_FRAME] = 2;
@ -1087,7 +1089,7 @@ void vp8_set_speed_features(VP8_COMP *cpi)
if (Speed >= 15)
sf->half_pixel_search = 0;
vpx_memset(cpi->mb.error_bins, 0, sizeof(cpi->mb.error_bins));
memset(cpi->mb.error_bins, 0, sizeof(cpi->mb.error_bins));
}; /* switch */
@ -1298,7 +1300,7 @@ void vp8_alloc_compressor_data(VP8_COMP *cpi)
CHECK_MEM_ERROR(cpi->active_map,
vpx_calloc(cm->mb_rows * cm->mb_cols,
sizeof(*cpi->active_map)));
vpx_memset(cpi->active_map , 1, (cm->mb_rows * cm->mb_cols));
memset(cpi->active_map , 1, (cm->mb_rows * cm->mb_cols));
#if CONFIG_MULTITHREAD
if (width < 640)
@ -1891,7 +1893,7 @@ struct VP8_COMP* vp8_create_compressor(VP8_CONFIG *oxcf)
cm = &cpi->common;
vpx_memset(cpi, 0, sizeof(VP8_COMP));
memset(cpi, 0, sizeof(VP8_COMP));
if (setjmp(cm->error.jmp))
{
@ -2010,6 +2012,8 @@ struct VP8_COMP* vp8_create_compressor(VP8_CONFIG *oxcf)
cpi->source_alt_ref_active = 0;
cpi->common.refresh_alt_ref_frame = 0;
cpi->force_maxqp = 0;
cpi->b_calculate_psnr = CONFIG_INTERNAL_STATS;
#if CONFIG_INTERNAL_STATS
cpi->b_calculate_ssimg = 0;
@ -2126,55 +2130,55 @@ struct VP8_COMP* vp8_create_compressor(VP8_CONFIG *oxcf)
}
#endif
cpi->fn_ptr[BLOCK_16X16].sdf = vp8_sad16x16;
cpi->fn_ptr[BLOCK_16X16].vf = vp8_variance16x16;
cpi->fn_ptr[BLOCK_16X16].sdf = vpx_sad16x16;
cpi->fn_ptr[BLOCK_16X16].vf = vpx_variance16x16;
cpi->fn_ptr[BLOCK_16X16].svf = vp8_sub_pixel_variance16x16;
cpi->fn_ptr[BLOCK_16X16].svf_halfpix_h = vp8_variance_halfpixvar16x16_h;
cpi->fn_ptr[BLOCK_16X16].svf_halfpix_v = vp8_variance_halfpixvar16x16_v;
cpi->fn_ptr[BLOCK_16X16].svf_halfpix_hv = vp8_variance_halfpixvar16x16_hv;
cpi->fn_ptr[BLOCK_16X16].sdx3f = vp8_sad16x16x3;
cpi->fn_ptr[BLOCK_16X16].sdx8f = vp8_sad16x16x8;
cpi->fn_ptr[BLOCK_16X16].sdx4df = vp8_sad16x16x4d;
cpi->fn_ptr[BLOCK_16X16].sdx3f = vpx_sad16x16x3;
cpi->fn_ptr[BLOCK_16X16].sdx8f = vpx_sad16x16x8;
cpi->fn_ptr[BLOCK_16X16].sdx4df = vpx_sad16x16x4d;
cpi->fn_ptr[BLOCK_16X8].sdf = vp8_sad16x8;
cpi->fn_ptr[BLOCK_16X8].vf = vp8_variance16x8;
cpi->fn_ptr[BLOCK_16X8].sdf = vpx_sad16x8;
cpi->fn_ptr[BLOCK_16X8].vf = vpx_variance16x8;
cpi->fn_ptr[BLOCK_16X8].svf = vp8_sub_pixel_variance16x8;
cpi->fn_ptr[BLOCK_16X8].svf_halfpix_h = NULL;
cpi->fn_ptr[BLOCK_16X8].svf_halfpix_v = NULL;
cpi->fn_ptr[BLOCK_16X8].svf_halfpix_hv = NULL;
cpi->fn_ptr[BLOCK_16X8].sdx3f = vp8_sad16x8x3;
cpi->fn_ptr[BLOCK_16X8].sdx8f = vp8_sad16x8x8;
cpi->fn_ptr[BLOCK_16X8].sdx4df = vp8_sad16x8x4d;
cpi->fn_ptr[BLOCK_16X8].sdx3f = vpx_sad16x8x3;
cpi->fn_ptr[BLOCK_16X8].sdx8f = vpx_sad16x8x8;
cpi->fn_ptr[BLOCK_16X8].sdx4df = vpx_sad16x8x4d;
cpi->fn_ptr[BLOCK_8X16].sdf = vp8_sad8x16;
cpi->fn_ptr[BLOCK_8X16].vf = vp8_variance8x16;
cpi->fn_ptr[BLOCK_8X16].sdf = vpx_sad8x16;
cpi->fn_ptr[BLOCK_8X16].vf = vpx_variance8x16;
cpi->fn_ptr[BLOCK_8X16].svf = vp8_sub_pixel_variance8x16;
cpi->fn_ptr[BLOCK_8X16].svf_halfpix_h = NULL;
cpi->fn_ptr[BLOCK_8X16].svf_halfpix_v = NULL;
cpi->fn_ptr[BLOCK_8X16].svf_halfpix_hv = NULL;
cpi->fn_ptr[BLOCK_8X16].sdx3f = vp8_sad8x16x3;
cpi->fn_ptr[BLOCK_8X16].sdx8f = vp8_sad8x16x8;
cpi->fn_ptr[BLOCK_8X16].sdx4df = vp8_sad8x16x4d;
cpi->fn_ptr[BLOCK_8X16].sdx3f = vpx_sad8x16x3;
cpi->fn_ptr[BLOCK_8X16].sdx8f = vpx_sad8x16x8;
cpi->fn_ptr[BLOCK_8X16].sdx4df = vpx_sad8x16x4d;
cpi->fn_ptr[BLOCK_8X8].sdf = vp8_sad8x8;
cpi->fn_ptr[BLOCK_8X8].vf = vp8_variance8x8;
cpi->fn_ptr[BLOCK_8X8].sdf = vpx_sad8x8;
cpi->fn_ptr[BLOCK_8X8].vf = vpx_variance8x8;
cpi->fn_ptr[BLOCK_8X8].svf = vp8_sub_pixel_variance8x8;
cpi->fn_ptr[BLOCK_8X8].svf_halfpix_h = NULL;
cpi->fn_ptr[BLOCK_8X8].svf_halfpix_v = NULL;
cpi->fn_ptr[BLOCK_8X8].svf_halfpix_hv = NULL;
cpi->fn_ptr[BLOCK_8X8].sdx3f = vp8_sad8x8x3;
cpi->fn_ptr[BLOCK_8X8].sdx8f = vp8_sad8x8x8;
cpi->fn_ptr[BLOCK_8X8].sdx4df = vp8_sad8x8x4d;
cpi->fn_ptr[BLOCK_8X8].sdx3f = vpx_sad8x8x3;
cpi->fn_ptr[BLOCK_8X8].sdx8f = vpx_sad8x8x8;
cpi->fn_ptr[BLOCK_8X8].sdx4df = vpx_sad8x8x4d;
cpi->fn_ptr[BLOCK_4X4].sdf = vp8_sad4x4;
cpi->fn_ptr[BLOCK_4X4].vf = vp8_variance4x4;
cpi->fn_ptr[BLOCK_4X4].sdf = vpx_sad4x4;
cpi->fn_ptr[BLOCK_4X4].vf = vpx_variance4x4;
cpi->fn_ptr[BLOCK_4X4].svf = vp8_sub_pixel_variance4x4;
cpi->fn_ptr[BLOCK_4X4].svf_halfpix_h = NULL;
cpi->fn_ptr[BLOCK_4X4].svf_halfpix_v = NULL;
cpi->fn_ptr[BLOCK_4X4].svf_halfpix_hv = NULL;
cpi->fn_ptr[BLOCK_4X4].sdx3f = vp8_sad4x4x3;
cpi->fn_ptr[BLOCK_4X4].sdx8f = vp8_sad4x4x8;
cpi->fn_ptr[BLOCK_4X4].sdx4df = vp8_sad4x4x4d;
cpi->fn_ptr[BLOCK_4X4].sdx3f = vpx_sad4x4x3;
cpi->fn_ptr[BLOCK_4X4].sdx8f = vpx_sad4x4x8;
cpi->fn_ptr[BLOCK_4X4].sdx4df = vpx_sad4x4x4d;
#if ARCH_X86 || ARCH_X86_64
cpi->fn_ptr[BLOCK_16X16].copymem = vp8_copy32xn;
@ -2554,7 +2558,7 @@ static uint64_t calc_plane_error(unsigned char *orig, int orig_stride,
{
unsigned int sse;
vp8_mse16x16(orig + col, orig_stride,
vpx_mse16x16(orig + col, orig_stride,
recon + col, recon_stride,
&sse);
total_sse += sse;
@ -2867,7 +2871,7 @@ static void update_alt_ref_frame_stats(VP8_COMP *cpi)
}
/* Update data structure that monitors level of reference to last GF */
vpx_memset(cpi->gf_active_flags, 1, (cm->mb_rows * cm->mb_cols));
memset(cpi->gf_active_flags, 1, (cm->mb_rows * cm->mb_cols));
cpi->gf_active_count = cm->mb_rows * cm->mb_cols;
/* this frame refreshes means next frames don't unless specified by user */
@ -2916,7 +2920,7 @@ static void update_golden_frame_stats(VP8_COMP *cpi)
}
/* Update data structure that monitors level of reference to last GF */
vpx_memset(cpi->gf_active_flags, 1, (cm->mb_rows * cm->mb_cols));
memset(cpi->gf_active_flags, 1, (cm->mb_rows * cm->mb_cols));
cpi->gf_active_count = cm->mb_rows * cm->mb_cols;
/* this frame refreshes means next frames don't unless specified by
@ -3380,7 +3384,7 @@ static int measure_square_diff_partial(YV12_BUFFER_CONFIG *source,
int index = block_index_row + (j >> 4);
if (cpi->consec_zero_last[index] >= min_consec_zero_last) {
unsigned int sse;
Total += vp8_mse16x16(src + j,
Total += vpx_mse16x16(src + j,
source->y_stride,
dst + j, dest->y_stride,
&sse);
@ -3444,7 +3448,7 @@ static void process_denoiser_mode_change(VP8_COMP *cpi) {
int index = block_index_row + (j >> 4);
if (cpi->consec_zero_last[index] >= min_consec_zero_last) {
unsigned int sse;
const unsigned int var = vp8_variance16x16(src + j,
const unsigned int var = vpx_variance16x16(src + j,
ystride,
dst + j,
ystride,
@ -3454,7 +3458,7 @@ static void process_denoiser_mode_change(VP8_COMP *cpi) {
// is small (to avoid effects from lighting change).
if ((sse - var) < 128) {
unsigned int sse2;
const unsigned int act = vp8_variance16x16(src + j,
const unsigned int act = vpx_variance16x16(src + j,
ystride,
const_source,
0,
@ -3830,9 +3834,9 @@ static void encode_frame_to_data_rate
}
// Reset the zero_last counter to 0 on key frame.
vpx_memset(cpi->consec_zero_last, 0, cm->mb_rows * cm->mb_cols);
vpx_memset(cpi->consec_zero_last_mvbias, 0,
(cpi->common.mb_rows * cpi->common.mb_cols));
memset(cpi->consec_zero_last, 0, cm->mb_rows * cm->mb_cols);
memset(cpi->consec_zero_last_mvbias, 0,
(cpi->common.mb_rows * cpi->common.mb_cols));
}
#if 0
@ -4183,7 +4187,10 @@ static void encode_frame_to_data_rate
*/
if (cpi->cyclic_refresh_mode_enabled)
{
if (cpi->current_layer==0)
// Special case for screen_content_mode with golden frame updates.
int disable_cr_gf = (cpi->oxcf.screen_content_mode == 2 &&
cm->refresh_golden_frame);
if (cpi->current_layer == 0 && cpi->force_maxqp == 0 && !disable_cr_gf)
cyclic_background_refresh(cpi, Q, 0);
else
disable_segmentation(cpi);
@ -4362,9 +4369,9 @@ static void encode_frame_to_data_rate
disable_segmentation(cpi);
}
// Reset the zero_last counter to 0 on key frame.
vpx_memset(cpi->consec_zero_last, 0, cm->mb_rows * cm->mb_cols);
vpx_memset(cpi->consec_zero_last_mvbias, 0,
(cpi->common.mb_rows * cpi->common.mb_cols));
memset(cpi->consec_zero_last, 0, cm->mb_rows * cm->mb_cols);
memset(cpi->consec_zero_last_mvbias, 0,
(cpi->common.mb_rows * cpi->common.mb_cols));
vp8_set_quantizer(cpi, Q);
}
@ -4387,7 +4394,7 @@ static void encode_frame_to_data_rate
if (cm->refresh_entropy_probs == 0)
{
/* save a copy for later refresh */
vpx_memcpy(&cm->lfc, &cm->fc, sizeof(cm->fc));
memcpy(&cm->lfc, &cm->fc, sizeof(cm->fc));
}
vp8_update_coef_context(cpi);
@ -4405,6 +4412,11 @@ static void encode_frame_to_data_rate
/* transform / motion compensation build reconstruction frame */
vp8_encode_frame(cpi);
if (cpi->oxcf.screen_content_mode == 2) {
if (vp8_drop_encodedframe_overshoot(cpi, Q))
return;
}
cpi->projected_frame_size -= vp8_estimate_entropy_savings(cpi);
cpi->projected_frame_size = (cpi->projected_frame_size > 0) ? cpi->projected_frame_size : 0;
#endif
@ -5613,19 +5625,19 @@ int vp8_get_compressed_data(VP8_COMP *cpi, unsigned int *frame_flags, unsigned l
if (cm->refresh_entropy_probs == 0)
{
vpx_memcpy(&cm->fc, &cm->lfc, sizeof(cm->fc));
memcpy(&cm->fc, &cm->lfc, sizeof(cm->fc));
}
/* Save the contexts separately for alt ref, gold and last. */
/* (TODO jbb -> Optimize this with pointers to avoid extra copies. ) */
if(cm->refresh_alt_ref_frame)
vpx_memcpy(&cpi->lfc_a, &cm->fc, sizeof(cm->fc));
memcpy(&cpi->lfc_a, &cm->fc, sizeof(cm->fc));
if(cm->refresh_golden_frame)
vpx_memcpy(&cpi->lfc_g, &cm->fc, sizeof(cm->fc));
memcpy(&cpi->lfc_g, &cm->fc, sizeof(cm->fc));
if(cm->refresh_last_frame)
vpx_memcpy(&cpi->lfc_n, &cm->fc, sizeof(cm->fc));
memcpy(&cpi->lfc_n, &cm->fc, sizeof(cm->fc));
/* if its a dropped frame honor the requests on subsequent frames */
if (*size > 0)
@ -5934,7 +5946,7 @@ int vp8_set_active_map(VP8_COMP *cpi, unsigned char *map, unsigned int rows, uns
{
if (map)
{
vpx_memcpy(cpi->active_map, map, rows * cols);
memcpy(cpi->active_map, map, rows * cols);
cpi->active_map_enabled = 1;
}
else
@ -5981,7 +5993,8 @@ int vp8_calc_ss_err(YV12_BUFFER_CONFIG *source, YV12_BUFFER_CONFIG *dest)
for (j = 0; j < source->y_width; j += 16)
{
unsigned int sse;
Total += vp8_mse16x16(src + j, source->y_stride, dst + j, dest->y_stride, &sse);
Total += vpx_mse16x16(src + j, source->y_stride,
dst + j, dest->y_stride, &sse);
}
src += 16 * source->y_stride;

7
media/libvpx/vp8/encoder/onyx_int.h
Просмотреть файл

@ -526,6 +526,8 @@ typedef struct VP8_COMP
// Measure of average squared difference between source and denoised signal.
int mse_source_denoised;
int force_maxqp;
#if CONFIG_MULTITHREAD
/* multithread data */
int * mt_current_mb_col;
@ -714,6 +716,11 @@ typedef struct VP8_COMP
} rd_costs;
} VP8_COMP;
void vp8_alloc_compressor_data(VP8_COMP *cpi);
int vp8_reverse_trans(int x);
void vp8_new_framerate(VP8_COMP *cpi, double framerate);
void vp8_loopfilter_frame(VP8_COMP *cpi, VP8_COMMON *cm);
void vp8_pack_bitstream(VP8_COMP *cpi, unsigned char *dest,
unsigned char *dest_end, unsigned long *size);

56
media/libvpx/vp8/encoder/pickinter.c
Просмотреть файл

@ -11,6 +11,7 @@
#include <limits.h>
#include "vpx_config.h"
#include "./vpx_dsp_rtcd.h"
#include "onyx_int.h"
#include "modecosts.h"
#include "encodeintra.h"
@ -29,8 +30,6 @@
#include "denoising.h"
#endif
extern int VP8_UVSSE(MACROBLOCK *x);
#ifdef SPEEDSTATS
extern unsigned int cnt_pm;
#endif
@ -38,8 +37,6 @@ extern unsigned int cnt_pm;
extern const int vp8_ref_frame_order[MAX_MODES];
extern const MB_PREDICTION_MODE vp8_mode_order[MAX_MODES];
extern int vp8_cost_mv_ref(MB_PREDICTION_MODE m, const int near_mv_ref_ct[4]);
// Fixed point implementation of a skin color classifier. Skin color
// is model by a Gaussian distribution in the CbCr color space.
// See ../../test/skin_color_detector_test.cc where the reference
@ -219,33 +216,6 @@ int vp8_get_inter_mbpred_error(MACROBLOCK *mb,
}
unsigned int vp8_get4x4sse_cs_c
(
const unsigned char *src_ptr,
int source_stride,
const unsigned char *ref_ptr,
int recon_stride
)
{
int distortion = 0;
int r, c;
for (r = 0; r < 4; r++)
{
for (c = 0; c < 4; c++)
{
int diff = src_ptr[c] - ref_ptr[c];
distortion += diff * diff;
}
src_ptr += source_stride;
ref_ptr += recon_stride;
}
return distortion;
}
static int get_prediction_error(BLOCK *be, BLOCKD *b)
{
unsigned char *sptr;
@ -253,7 +223,7 @@ static int get_prediction_error(BLOCK *be, BLOCKD *b)
sptr = (*(be->base_src) + be->src);
dptr = b->predictor;
return vp8_get4x4sse_cs(sptr, be->src_stride, dptr, 16);
return vpx_get4x4sse_cs(sptr, be->src_stride, dptr, 16);
}
@ -862,8 +832,8 @@ void vp8_pick_inter_mode(VP8_COMP *cpi, MACROBLOCK *x, int recon_yoffset,
mode_mv = mode_mv_sb[sign_bias];
best_ref_mv.as_int = 0;
vpx_memset(mode_mv_sb, 0, sizeof(mode_mv_sb));
vpx_memset(&best_mbmode, 0, sizeof(best_mbmode));
memset(mode_mv_sb, 0, sizeof(mode_mv_sb));
memset(&best_mbmode, 0, sizeof(best_mbmode));
/* Setup search priorities */
#if CONFIG_MULTI_RES_ENCODING
@ -1041,7 +1011,7 @@ void vp8_pick_inter_mode(VP8_COMP *cpi, MACROBLOCK *x, int recon_yoffset,
else
{
rate2 += rate;
distortion2 = vp8_variance16x16(
distortion2 = vpx_variance16x16(
*(b->base_src), b->src_stride,
x->e_mbd.predictor, 16, &sse);
this_rd = RDCOST(x->rdmult, x->rddiv, rate2, distortion2);
@ -1070,7 +1040,7 @@ void vp8_pick_inter_mode(VP8_COMP *cpi, MACROBLOCK *x, int recon_yoffset,
xd->dst.y_stride,
xd->predictor,
16);
distortion2 = vp8_variance16x16
distortion2 = vpx_variance16x16
(*(b->base_src), b->src_stride,
x->e_mbd.predictor, 16, &sse);
rate2 += x->mbmode_cost[x->e_mbd.frame_type][x->e_mbd.mode_info_context->mbmi.mode];
@ -1348,8 +1318,8 @@ void vp8_pick_inter_mode(VP8_COMP *cpi, MACROBLOCK *x, int recon_yoffset,
*returndistortion = distortion2;
best_rd_sse = sse;
best_rd = this_rd;
vpx_memcpy(&best_mbmode, &x->e_mbd.mode_info_context->mbmi,
sizeof(MB_MODE_INFO));
memcpy(&best_mbmode, &x->e_mbd.mode_info_context->mbmi,
sizeof(MB_MODE_INFO));
/* Testing this mode gave rise to an improvement in best error
* score. Lower threshold a bit for next time
@ -1487,8 +1457,8 @@ void vp8_pick_inter_mode(VP8_COMP *cpi, MACROBLOCK *x, int recon_yoffset,
if (this_rd < best_rd)
{
vpx_memcpy(&best_mbmode, &x->e_mbd.mode_info_context->mbmi,
sizeof(MB_MODE_INFO));
memcpy(&best_mbmode, &x->e_mbd.mode_info_context->mbmi,
sizeof(MB_MODE_INFO));
}
}
@ -1512,8 +1482,8 @@ void vp8_pick_inter_mode(VP8_COMP *cpi, MACROBLOCK *x, int recon_yoffset,
/* set to the best mb mode, this copy can be skip if x->skip since it
* already has the right content */
if (!x->skip)
vpx_memcpy(&x->e_mbd.mode_info_context->mbmi, &best_mbmode,
sizeof(MB_MODE_INFO));
memcpy(&x->e_mbd.mode_info_context->mbmi, &best_mbmode,
sizeof(MB_MODE_INFO));
if (best_mbmode.mode <= B_PRED)
{
@ -1551,7 +1521,7 @@ void vp8_pick_intra_mode(MACROBLOCK *x, int *rate_)
xd->dst.y_stride,
xd->predictor,
16);
distortion = vp8_variance16x16
distortion = vpx_variance16x16
(*(b->base_src), b->src_stride, xd->predictor, 16, &sse);
rate = x->mbmode_cost[xd->frame_type][mode];
this_rd = RDCOST(x->rdmult, x->rddiv, rate, distortion);

12
media/libvpx/vp8/encoder/picklpf.c
Просмотреть файл

@ -9,6 +9,7 @@
*/
#include "./vpx_dsp_rtcd.h"
#include "./vpx_scale_rtcd.h"
#include "vp8/common/onyxc_int.h"
#include "onyx_int.h"
@ -49,7 +50,7 @@ static void yv12_copy_partial_frame(YV12_BUFFER_CONFIG *src_ybc,
src_y = src_ybc->y_buffer + yoffset;
dst_y = dst_ybc->y_buffer + yoffset;
vpx_memcpy(dst_y, src_y, ystride * linestocopy);
memcpy(dst_y, src_y, ystride * linestocopy);
}
static int calc_partial_ssl_err(YV12_BUFFER_CONFIG *source,
@ -83,7 +84,7 @@ static int calc_partial_ssl_err(YV12_BUFFER_CONFIG *source,
for (j = 0; j < source->y_width; j += 16)
{
unsigned int sse;
Total += vp8_mse16x16(src + j, source->y_stride,
Total += vpx_mse16x16(src + j, source->y_stride,
dst + j, dest->y_stride,
&sse);
}
@ -142,7 +143,7 @@ void vp8cx_pick_filter_level_fast(YV12_BUFFER_CONFIG *sd, VP8_COMP *cpi)
int min_filter_level = get_min_filter_level(cpi, cm->base_qindex);
int max_filter_level = get_max_filter_level(cpi, cm->base_qindex);
int filt_val;
int best_filt_val = cm->filter_level;
int best_filt_val;
YV12_BUFFER_CONFIG * saved_frame = cm->frame_to_show;
/* Replace unfiltered frame buffer with a new one */
@ -274,8 +275,7 @@ void vp8cx_pick_filter_level(YV12_BUFFER_CONFIG *sd, VP8_COMP *cpi)
int filter_step;
int filt_high = 0;
/* Start search at previous frame filter level */
int filt_mid = cm->filter_level;
int filt_mid;
int filt_low = 0;
int filt_best;
int filt_direction = 0;
@ -287,7 +287,7 @@ void vp8cx_pick_filter_level(YV12_BUFFER_CONFIG *sd, VP8_COMP *cpi)
YV12_BUFFER_CONFIG * saved_frame = cm->frame_to_show;
vpx_memset(ss_err, 0, sizeof(ss_err));
memset(ss_err, 0, sizeof(ss_err));
/* Replace unfiltered frame buffer with a new one */
cm->frame_to_show = &cpi->pick_lf_lvl_frame;

4
media/libvpx/vp8/encoder/quantize.c
Просмотреть файл

@ -65,8 +65,8 @@ void vp8_regular_quantize_b_c(BLOCK *b, BLOCKD *d)
short *dequant_ptr = d->dequant;
short zbin_oq_value = b->zbin_extra;
vpx_memset(qcoeff_ptr, 0, 32);
vpx_memset(dqcoeff_ptr, 0, 32);
memset(qcoeff_ptr, 0, 32);
memset(dqcoeff_ptr, 0, 32);
eob = -1;

56
media/libvpx/vp8/encoder/ratectrl.c
Просмотреть файл

@ -296,7 +296,7 @@ void vp8_setup_key_frame(VP8_COMP *cpi)
vp8_default_coef_probs(& cpi->common);
vpx_memcpy(cpi->common.fc.mvc, vp8_default_mv_context, sizeof(vp8_default_mv_context));
memcpy(cpi->common.fc.mvc, vp8_default_mv_context, sizeof(vp8_default_mv_context));
{
int flag[2] = {1, 1};
vp8_build_component_cost_table(cpi->mb.mvcost, (const MV_CONTEXT *) cpi->common.fc.mvc, flag);
@ -305,9 +305,9 @@ void vp8_setup_key_frame(VP8_COMP *cpi)
/* Make sure we initialize separate contexts for altref,gold, and normal.
* TODO shouldn't need 3 different copies of structure to do this!
*/
vpx_memcpy(&cpi->lfc_a, &cpi->common.fc, sizeof(cpi->common.fc));
vpx_memcpy(&cpi->lfc_g, &cpi->common.fc, sizeof(cpi->common.fc));
vpx_memcpy(&cpi->lfc_n, &cpi->common.fc, sizeof(cpi->common.fc));
memcpy(&cpi->lfc_a, &cpi->common.fc, sizeof(cpi->common.fc));
memcpy(&cpi->lfc_g, &cpi->common.fc, sizeof(cpi->common.fc));
memcpy(&cpi->lfc_n, &cpi->common.fc, sizeof(cpi->common.fc));
cpi->common.filter_level = cpi->common.base_qindex * 3 / 8 ;
@ -1215,6 +1215,11 @@ int vp8_regulate_q(VP8_COMP *cpi, int target_bits_per_frame)
{
int Q = cpi->active_worst_quality;
if (cpi->force_maxqp == 1) {
cpi->active_worst_quality = cpi->worst_quality;
return cpi->worst_quality;
}
/* Reset Zbin OQ value */
cpi->mb.zbin_over_quant = 0;
@ -1559,3 +1564,46 @@ int vp8_pick_frame_size(VP8_COMP *cpi)
}
return 1;
}
// If this just encoded frame (mcomp/transform/quant, but before loopfilter and
// pack_bitstream) has large overshoot, and was not being encoded close to the
// max QP, then drop this frame and force next frame to be encoded at max QP.
// Condition this on 1 pass CBR with screen content mode and frame dropper off.
// TODO(marpan): Should do this exit condition during the encode_frame
// (i.e., halfway during the encoding of the frame) to save cycles.
int vp8_drop_encodedframe_overshoot(VP8_COMP *cpi, int Q) {
if (cpi->pass == 0 &&
cpi->oxcf.end_usage == USAGE_STREAM_FROM_SERVER &&
cpi->drop_frames_allowed == 0 &&
cpi->common.frame_type != KEY_FRAME) {
// Note: the "projected_frame_size" from encode_frame() only gives estimate
// of mode/motion vector rate (in non-rd mode): so below we only require
// that projected_frame_size is somewhat greater than per-frame-bandwidth,
// but add additional condition with high threshold on prediction residual.
// QP threshold: only allow dropping if we are not close to qp_max.
int thresh_qp = 3 * cpi->worst_quality >> 2;
// Rate threshold, in bytes.
int thresh_rate = 2 * (cpi->av_per_frame_bandwidth >> 3);
// Threshold for the average (over all macroblocks) of the pixel-sum
// residual error over 16x16 block. Should add QP dependence on threshold?
int thresh_pred_err_mb = (256 << 4);
int pred_err_mb = (int)(cpi->mb.prediction_error / cpi->common.MBs);
if (Q < thresh_qp &&
cpi->projected_frame_size > thresh_rate &&
pred_err_mb > thresh_pred_err_mb) {
// Drop this frame: advance frame counters, and set force_maxqp flag.
cpi->common.current_video_frame++;
cpi->frames_since_key++;
// Flag to indicate we will force next frame to be encoded at max QP.
cpi->force_maxqp = 1;
return 1;
} else {
cpi->force_maxqp = 0;
return 0;
}
cpi->force_maxqp = 0;
return 0;
}
cpi->force_maxqp = 0;
return 0;
}

2
media/libvpx/vp8/encoder/ratectrl.h
Просмотреть файл

@ -30,6 +30,8 @@ extern void vp8_compute_frame_size_bounds(VP8_COMP *cpi, int *frame_under_shoot_
/* return of 0 means drop frame */
extern int vp8_pick_frame_size(VP8_COMP *cpi);
extern int vp8_drop_encodedframe_overshoot(VP8_COMP *cpi, int Q);
#ifdef __cplusplus
} // extern "C"
#endif

81
media/libvpx/vp8/encoder/rdopt.c
Просмотреть файл

@ -15,6 +15,7 @@
#include <assert.h>
#include "vpx_config.h"
#include "vp8_rtcd.h"
#include "./vpx_dsp_rtcd.h"
#include "tokenize.h"
#include "treewriter.h"
#include "onyx_int.h"
@ -507,9 +508,9 @@ int VP8_UVSSE(MACROBLOCK *x)
}
else
{
vp8_variance8x8(uptr, pre_stride,
vpx_variance8x8(uptr, pre_stride,
upred_ptr, uv_stride, &sse2);
vp8_variance8x8(vptr, pre_stride,
vpx_variance8x8(vptr, pre_stride,
vpred_ptr, uv_stride, &sse1);
sse2 += sse1;
}
@ -555,8 +556,8 @@ static int vp8_rdcost_mby(MACROBLOCK *mb)
ENTROPY_CONTEXT *ta;
ENTROPY_CONTEXT *tl;
vpx_memcpy(&t_above, mb->e_mbd.above_context, sizeof(ENTROPY_CONTEXT_PLANES));
vpx_memcpy(&t_left, mb->e_mbd.left_context, sizeof(ENTROPY_CONTEXT_PLANES));
memcpy(&t_above, mb->e_mbd.above_context, sizeof(ENTROPY_CONTEXT_PLANES));
memcpy(&t_left, mb->e_mbd.left_context, sizeof(ENTROPY_CONTEXT_PLANES));
ta = (ENTROPY_CONTEXT *)&t_above;
tl = (ENTROPY_CONTEXT *)&t_left;
@ -650,8 +651,8 @@ static int rd_pick_intra4x4block(
* a temp buffer that meets the stride requirements, but we are only
* interested in the left 4x4 block
* */
DECLARE_ALIGNED_ARRAY(16, unsigned char, best_predictor, 16*4);
DECLARE_ALIGNED_ARRAY(16, short, best_dqcoeff, 16);
DECLARE_ALIGNED(16, unsigned char, best_predictor[16*4]);
DECLARE_ALIGNED(16, short, best_dqcoeff[16]);
int dst_stride = x->e_mbd.dst.y_stride;
unsigned char *dst = x->e_mbd.dst.y_buffer + b->offset;
@ -691,7 +692,7 @@ static int rd_pick_intra4x4block(
*a = tempa;
*l = templ;
copy_predictor(best_predictor, b->predictor);
vpx_memcpy(best_dqcoeff, b->dqcoeff, 32);
memcpy(best_dqcoeff, b->dqcoeff, 32);
}
}
b->bmi.as_mode = *best_mode;
@ -715,8 +716,8 @@ static int rd_pick_intra4x4mby_modes(MACROBLOCK *mb, int *Rate,
ENTROPY_CONTEXT *tl;
const int *bmode_costs;
vpx_memcpy(&t_above, mb->e_mbd.above_context, sizeof(ENTROPY_CONTEXT_PLANES));
vpx_memcpy(&t_left, mb->e_mbd.left_context, sizeof(ENTROPY_CONTEXT_PLANES));
memcpy(&t_above, mb->e_mbd.above_context, sizeof(ENTROPY_CONTEXT_PLANES));
memcpy(&t_left, mb->e_mbd.left_context, sizeof(ENTROPY_CONTEXT_PLANES));
ta = (ENTROPY_CONTEXT *)&t_above;
tl = (ENTROPY_CONTEXT *)&t_left;
@ -820,8 +821,8 @@ static int rd_cost_mbuv(MACROBLOCK *mb)
ENTROPY_CONTEXT *ta;
ENTROPY_CONTEXT *tl;
vpx_memcpy(&t_above, mb->e_mbd.above_context, sizeof(ENTROPY_CONTEXT_PLANES));
vpx_memcpy(&t_left, mb->e_mbd.left_context, sizeof(ENTROPY_CONTEXT_PLANES));
memcpy(&t_above, mb->e_mbd.above_context, sizeof(ENTROPY_CONTEXT_PLANES));
memcpy(&t_left, mb->e_mbd.left_context, sizeof(ENTROPY_CONTEXT_PLANES));
ta = (ENTROPY_CONTEXT *)&t_above;
tl = (ENTROPY_CONTEXT *)&t_left;
@ -1128,8 +1129,8 @@ static void rd_check_segment(VP8_COMP *cpi, MACROBLOCK *x,
ENTROPY_CONTEXT *ta_b;
ENTROPY_CONTEXT *tl_b;
vpx_memcpy(&t_above, x->e_mbd.above_context, sizeof(ENTROPY_CONTEXT_PLANES));
vpx_memcpy(&t_left, x->e_mbd.left_context, sizeof(ENTROPY_CONTEXT_PLANES));
memcpy(&t_above, x->e_mbd.above_context, sizeof(ENTROPY_CONTEXT_PLANES));
memcpy(&t_left, x->e_mbd.left_context, sizeof(ENTROPY_CONTEXT_PLANES));
ta = (ENTROPY_CONTEXT *)&t_above;
tl = (ENTROPY_CONTEXT *)&t_left;
@ -1172,8 +1173,8 @@ static void rd_check_segment(VP8_COMP *cpi, MACROBLOCK *x,
ENTROPY_CONTEXT *ta_s;
ENTROPY_CONTEXT *tl_s;
vpx_memcpy(&t_above_s, &t_above, sizeof(ENTROPY_CONTEXT_PLANES));
vpx_memcpy(&t_left_s, &t_left, sizeof(ENTROPY_CONTEXT_PLANES));
memcpy(&t_above_s, &t_above, sizeof(ENTROPY_CONTEXT_PLANES));
memcpy(&t_left_s, &t_left, sizeof(ENTROPY_CONTEXT_PLANES));
ta_s = (ENTROPY_CONTEXT *)&t_above_s;
tl_s = (ENTROPY_CONTEXT *)&t_left_s;
@ -1329,14 +1330,14 @@ static void rd_check_segment(VP8_COMP *cpi, MACROBLOCK *x,
mode_selected = this_mode;
best_label_rd = this_rd;
vpx_memcpy(ta_b, ta_s, sizeof(ENTROPY_CONTEXT_PLANES));
vpx_memcpy(tl_b, tl_s, sizeof(ENTROPY_CONTEXT_PLANES));
memcpy(ta_b, ta_s, sizeof(ENTROPY_CONTEXT_PLANES));
memcpy(tl_b, tl_s, sizeof(ENTROPY_CONTEXT_PLANES));
}
} /*for each 4x4 mode*/
vpx_memcpy(ta, ta_b, sizeof(ENTROPY_CONTEXT_PLANES));
vpx_memcpy(tl, tl_b, sizeof(ENTROPY_CONTEXT_PLANES));
memcpy(ta, ta_b, sizeof(ENTROPY_CONTEXT_PLANES));
memcpy(tl, tl_b, sizeof(ENTROPY_CONTEXT_PLANES));
labels2mode(x, labels, i, mode_selected, &mode_mv[mode_selected],
bsi->ref_mv, x->mvcost);
@ -1392,7 +1393,7 @@ static int vp8_rd_pick_best_mbsegmentation(VP8_COMP *cpi, MACROBLOCK *x,
int i;
BEST_SEG_INFO bsi;
vpx_memset(&bsi, 0, sizeof(bsi));
memset(&bsi, 0, sizeof(bsi));
bsi.segment_rd = best_rd;
bsi.ref_mv = best_ref_mv;
@ -1661,7 +1662,6 @@ void vp8_mv_pred
mv.as_mv.row = mvx[vcnt/2];
mv.as_mv.col = mvy[vcnt/2];
find = 1;
/* sr is set to 0 to allow calling function to decide the search
* range.
*/
@ -1691,16 +1691,16 @@ void vp8_cal_sad(VP8_COMP *cpi, MACROBLOCKD *xd, MACROBLOCK *x, int recon_yoffse
}else if(xd->mb_to_top_edge==0)
{ /* only has left MB for sad calculation. */
near_sad[0] = near_sad[2] = INT_MAX;
near_sad[1] = cpi->fn_ptr[BLOCK_16X16].sdf(src_y_ptr, b->src_stride, xd->dst.y_buffer - 16,xd->dst.y_stride, UINT_MAX);
near_sad[1] = cpi->fn_ptr[BLOCK_16X16].sdf(src_y_ptr, b->src_stride, xd->dst.y_buffer - 16,xd->dst.y_stride);
}else if(xd->mb_to_left_edge ==0)
{ /* only has left MB for sad calculation. */
near_sad[1] = near_sad[2] = INT_MAX;
near_sad[0] = cpi->fn_ptr[BLOCK_16X16].sdf(src_y_ptr, b->src_stride, xd->dst.y_buffer - xd->dst.y_stride *16,xd->dst.y_stride, UINT_MAX);
near_sad[0] = cpi->fn_ptr[BLOCK_16X16].sdf(src_y_ptr, b->src_stride, xd->dst.y_buffer - xd->dst.y_stride *16,xd->dst.y_stride);
}else
{
near_sad[0] = cpi->fn_ptr[BLOCK_16X16].sdf(src_y_ptr, b->src_stride, xd->dst.y_buffer - xd->dst.y_stride *16,xd->dst.y_stride, UINT_MAX);
near_sad[1] = cpi->fn_ptr[BLOCK_16X16].sdf(src_y_ptr, b->src_stride, xd->dst.y_buffer - 16,xd->dst.y_stride, UINT_MAX);
near_sad[2] = cpi->fn_ptr[BLOCK_16X16].sdf(src_y_ptr, b->src_stride, xd->dst.y_buffer - xd->dst.y_stride *16 -16,xd->dst.y_stride, UINT_MAX);
near_sad[0] = cpi->fn_ptr[BLOCK_16X16].sdf(src_y_ptr, b->src_stride, xd->dst.y_buffer - xd->dst.y_stride *16,xd->dst.y_stride);
near_sad[1] = cpi->fn_ptr[BLOCK_16X16].sdf(src_y_ptr, b->src_stride, xd->dst.y_buffer - 16,xd->dst.y_stride);
near_sad[2] = cpi->fn_ptr[BLOCK_16X16].sdf(src_y_ptr, b->src_stride, xd->dst.y_buffer - xd->dst.y_stride *16 -16,xd->dst.y_stride);
}
if(cpi->common.last_frame_type != KEY_FRAME)
@ -1715,14 +1715,14 @@ void vp8_cal_sad(VP8_COMP *cpi, MACROBLOCKD *xd, MACROBLOCK *x, int recon_yoffse
if(xd->mb_to_bottom_edge==0) near_sad[7] = INT_MAX;
if(near_sad[4] != INT_MAX)
near_sad[4] = cpi->fn_ptr[BLOCK_16X16].sdf(src_y_ptr, b->src_stride, pre_y_buffer - pre_y_stride *16, pre_y_stride, UINT_MAX);
near_sad[4] = cpi->fn_ptr[BLOCK_16X16].sdf(src_y_ptr, b->src_stride, pre_y_buffer - pre_y_stride *16, pre_y_stride);
if(near_sad[5] != INT_MAX)
near_sad[5] = cpi->fn_ptr[BLOCK_16X16].sdf(src_y_ptr, b->src_stride, pre_y_buffer - 16, pre_y_stride, UINT_MAX);
near_sad[3] = cpi->fn_ptr[BLOCK_16X16].sdf(src_y_ptr, b->src_stride, pre_y_buffer, pre_y_stride, UINT_MAX);
near_sad[5] = cpi->fn_ptr[BLOCK_16X16].sdf(src_y_ptr, b->src_stride, pre_y_buffer - 16, pre_y_stride);
near_sad[3] = cpi->fn_ptr[BLOCK_16X16].sdf(src_y_ptr, b->src_stride, pre_y_buffer, pre_y_stride);
if(near_sad[6] != INT_MAX)
near_sad[6] = cpi->fn_ptr[BLOCK_16X16].sdf(src_y_ptr, b->src_stride, pre_y_buffer + 16, pre_y_stride, UINT_MAX);
near_sad[6] = cpi->fn_ptr[BLOCK_16X16].sdf(src_y_ptr, b->src_stride, pre_y_buffer + 16, pre_y_stride);
if(near_sad[7] != INT_MAX)
near_sad[7] = cpi->fn_ptr[BLOCK_16X16].sdf(src_y_ptr, b->src_stride, pre_y_buffer + pre_y_stride *16, pre_y_stride, UINT_MAX);
near_sad[7] = cpi->fn_ptr[BLOCK_16X16].sdf(src_y_ptr, b->src_stride, pre_y_buffer + pre_y_stride *16, pre_y_stride);
}
if(cpi->common.last_frame_type != KEY_FRAME)
@ -1784,7 +1784,7 @@ static int evaluate_inter_mode_rd(int mdcounts[4],
if(threshold < x->encode_breakout)
threshold = x->encode_breakout;
var = vp8_variance16x16
var = vpx_variance16x16
(*(b->base_src), b->src_stride,
x->e_mbd.predictor, 16, &sse);
@ -1926,8 +1926,8 @@ static void update_best_mode(BEST_MODE* best_mode, int this_rd,
(rd->distortion2-rd->distortion_uv));
best_mode->rd = this_rd;
vpx_memcpy(&best_mode->mbmode, &x->e_mbd.mode_info_context->mbmi, sizeof(MB_MODE_INFO));
vpx_memcpy(&best_mode->partition, x->partition_info, sizeof(PARTITION_INFO));
memcpy(&best_mode->mbmode, &x->e_mbd.mode_info_context->mbmi, sizeof(MB_MODE_INFO));
memcpy(&best_mode->partition, x->partition_info, sizeof(PARTITION_INFO));
if ((this_mode == B_PRED) || (this_mode == SPLITMV))
{
@ -1989,9 +1989,9 @@ void vp8_rd_pick_inter_mode(VP8_COMP *cpi, MACROBLOCK *x, int recon_yoffset,
best_mode.rd = INT_MAX;
best_mode.yrd = INT_MAX;
best_mode.intra_rd = INT_MAX;
vpx_memset(mode_mv_sb, 0, sizeof(mode_mv_sb));
vpx_memset(&best_mode.mbmode, 0, sizeof(best_mode.mbmode));
vpx_memset(&best_mode.bmodes, 0, sizeof(best_mode.bmodes));
memset(mode_mv_sb, 0, sizeof(mode_mv_sb));
memset(&best_mode.mbmode, 0, sizeof(best_mode.mbmode));
memset(&best_mode.bmodes, 0, sizeof(best_mode.bmodes));
/* Setup search priorities */
get_reference_search_order(cpi, ref_frame_map);
@ -2293,7 +2293,6 @@ void vp8_rd_pick_inter_mode(VP8_COMP *cpi, MACROBLOCK *x, int recon_yoffset,
mode_mv[NEWMV].as_int = d->bmi.mv.as_int;
/* Further step/diamond searches as necessary */
n = 0;
further_steps = (cpi->sf.max_step_search_steps - 1) - step_param;
n = num00;
@ -2560,8 +2559,6 @@ void vp8_rd_pick_inter_mode(VP8_COMP *cpi, MACROBLOCK *x, int recon_yoffset,
intra_rd_penalty, cpi, x);
if (this_rd < best_mode.rd || x->skip)
{
/* Note index of best mode so far */
best_mode_index = mode_index;
*returnrate = rd.rate2;
*returndistortion = rd.distortion2;
update_best_mode(&best_mode, this_rd, &rd, other_cost, x);
@ -2586,7 +2583,7 @@ void vp8_rd_pick_inter_mode(VP8_COMP *cpi, MACROBLOCK *x, int recon_yoffset,
/* macroblock modes */
vpx_memcpy(&x->e_mbd.mode_info_context->mbmi, &best_mode.mbmode, sizeof(MB_MODE_INFO));
memcpy(&x->e_mbd.mode_info_context->mbmi, &best_mode.mbmode, sizeof(MB_MODE_INFO));
if (best_mode.mbmode.mode == B_PRED)
{
@ -2599,7 +2596,7 @@ void vp8_rd_pick_inter_mode(VP8_COMP *cpi, MACROBLOCK *x, int recon_yoffset,
for (i = 0; i < 16; i++)
xd->mode_info_context->bmi[i].mv.as_int = best_mode.bmodes[i].mv.as_int;
vpx_memcpy(x->partition_info, &best_mode.partition, sizeof(PARTITION_INFO));
memcpy(x->partition_info, &best_mode.partition, sizeof(PARTITION_INFO));
x->e_mbd.mode_info_context->mbmi.mv.as_int =
x->partition_info->bmi[15].mv.as_int;

3
media/libvpx/vp8/encoder/rdopt.h
Просмотреть файл

@ -136,6 +136,9 @@ extern void vp8_mv_pred
int near_sadidx[]
);
void vp8_cal_sad(VP8_COMP *cpi, MACROBLOCKD *xd, MACROBLOCK *x, int recon_yoffset, int near_sadidx[]);
int VP8_UVSSE(MACROBLOCK *x);
int vp8_cost_mv_ref(MB_PREDICTION_MODE m, const int near_mv_ref_ct[4]);
void vp8_set_mbmode_and_mvs(MACROBLOCK *x, MB_PREDICTION_MODE mb, int_mv *mv);
#ifdef __cplusplus
} // extern "C"

2
media/libvpx/vp8/encoder/segmentation.c
Просмотреть файл

@ -23,7 +23,7 @@ void vp8_update_gf_useage_maps(VP8_COMP *cpi, VP8_COMMON *cm, MACROBLOCK *x)
if ((cm->frame_type == KEY_FRAME) || (cm->refresh_golden_frame))
{
/* Reset Gf useage monitors */
vpx_memset(cpi->gf_active_flags, 1, (cm->mb_rows * cm->mb_cols));
memset(cpi->gf_active_flags, 1, (cm->mb_rows * cm->mb_cols));
cpi->gf_active_count = cm->mb_rows * cm->mb_cols;
}
else

12
media/libvpx/vp8/encoder/temporal_filter.c
Просмотреть файл

@ -238,12 +238,12 @@ static void vp8_temporal_filter_iterate_c
int mb_rows = cpi->common.mb_rows;
int mb_y_offset = 0;
int mb_uv_offset = 0;
DECLARE_ALIGNED_ARRAY(16, unsigned int, accumulator, 16*16 + 8*8 + 8*8);
DECLARE_ALIGNED_ARRAY(16, unsigned short, count, 16*16 + 8*8 + 8*8);
DECLARE_ALIGNED(16, unsigned int, accumulator[16*16 + 8*8 + 8*8]);
DECLARE_ALIGNED(16, unsigned short, count[16*16 + 8*8 + 8*8]);
MACROBLOCKD *mbd = &cpi->mb.e_mbd;
YV12_BUFFER_CONFIG *f = cpi->frames[alt_ref_index];
unsigned char *dst1, *dst2;
DECLARE_ALIGNED_ARRAY(16, unsigned char, predictor, 16*16 + 8*8 + 8*8);
DECLARE_ALIGNED(16, unsigned char, predictor[16*16 + 8*8 + 8*8]);
/* Save input state */
unsigned char *y_buffer = mbd->pre.y_buffer;
@ -274,8 +274,8 @@ static void vp8_temporal_filter_iterate_c
int i, j, k;
int stride;
vpx_memset(accumulator, 0, 384*sizeof(unsigned int));
vpx_memset(count, 0, 384*sizeof(unsigned short));
memset(accumulator, 0, 384*sizeof(unsigned int));
memset(count, 0, 384*sizeof(unsigned short));
#if ALT_REF_MC_ENABLED
cpi->mb.mv_col_min = -((mb_col * 16) + (16 - 5));
@ -502,7 +502,7 @@ void vp8_temporal_filter_prepare_c
start_frame = distance + frames_to_blur_forward;
/* Setup frame pointers, NULL indicates frame not included in filter */
vpx_memset(cpi->frames, 0, max_frames*sizeof(YV12_BUFFER_CONFIG *));
memset(cpi->frames, 0, max_frames*sizeof(YV12_BUFFER_CONFIG *));
for (frame = 0; frame < frames_to_blur; frame++)
{
int which_buffer = start_frame - frame;

10
media/libvpx/vp8/encoder/tokenize.c
Просмотреть файл

@ -421,7 +421,7 @@ void vp8_tokenize_mb(VP8_COMP *cpi, MACROBLOCK *x, TOKENEXTRA **t)
void init_context_counters(void)
{
vpx_memset(context_counters, 0, sizeof(context_counters));
memset(context_counters, 0, sizeof(context_counters));
}
void print_context_counters()
@ -596,13 +596,13 @@ void vp8_fix_contexts(MACROBLOCKD *x)
/* Clear entropy contexts for Y2 blocks */
if (x->mode_info_context->mbmi.mode != B_PRED && x->mode_info_context->mbmi.mode != SPLITMV)
{
vpx_memset(x->above_context, 0, sizeof(ENTROPY_CONTEXT_PLANES));
vpx_memset(x->left_context, 0, sizeof(ENTROPY_CONTEXT_PLANES));
memset(x->above_context, 0, sizeof(ENTROPY_CONTEXT_PLANES));
memset(x->left_context, 0, sizeof(ENTROPY_CONTEXT_PLANES));
}
else
{
vpx_memset(x->above_context, 0, sizeof(ENTROPY_CONTEXT_PLANES)-1);
vpx_memset(x->left_context, 0, sizeof(ENTROPY_CONTEXT_PLANES)-1);
memset(x->above_context, 0, sizeof(ENTROPY_CONTEXT_PLANES)-1);
memset(x->left_context, 0, sizeof(ENTROPY_CONTEXT_PLANES)-1);
}
}

8
media/libvpx/vp8/encoder/x86/quantize_sse2.c
Просмотреть файл

@ -35,10 +35,10 @@
void vp8_regular_quantize_b_sse2(BLOCK *b, BLOCKD *d)
{
char eob = 0;
short *zbin_boost_ptr = b->zrun_zbin_boost;
short *zbin_boost_ptr;
short *qcoeff_ptr = d->qcoeff;
DECLARE_ALIGNED_ARRAY(16, short, x, 16);
DECLARE_ALIGNED_ARRAY(16, short, y, 16);
DECLARE_ALIGNED(16, short, x[16]);
DECLARE_ALIGNED(16, short, y[16]);
__m128i sz0, x0, sz1, x1, y0, y1, x_minus_zbin0, x_minus_zbin1;
__m128i quant_shift0 = _mm_load_si128((__m128i *)(b->quant_shift));
@ -55,7 +55,7 @@ void vp8_regular_quantize_b_sse2(BLOCK *b, BLOCKD *d)
__m128i dequant0 = _mm_load_si128((__m128i *)(d->dequant));
__m128i dequant1 = _mm_load_si128((__m128i *)(d->dequant + 8));
vpx_memset(qcoeff_ptr, 0, 32);
memset(qcoeff_ptr, 0, 32);
/* Duplicate to all lanes. */
zbin_extra = _mm_shufflelo_epi16(zbin_extra, 0);

19
media/libvpx/vp8/vp8_cx_iface.c
Просмотреть файл

@ -10,7 +10,9 @@
#include "./vpx_config.h"
#include "vp8_rtcd.h"
#include "./vp8_rtcd.h"
#include "./vpx_dsp_rtcd.h"
#include "./vpx_scale_rtcd.h"
#include "vpx/vpx_codec.h"
#include "vpx/internal/vpx_codec_internal.h"
#include "vpx_version.h"
@ -133,7 +135,7 @@ static vpx_codec_err_t validate_config(vpx_codec_alg_priv_t *ctx,
RANGE_CHECK(cfg, g_w, 1, 16383); /* 14 bits available */
RANGE_CHECK(cfg, g_h, 1, 16383); /* 14 bits available */
RANGE_CHECK(cfg, g_timebase.den, 1, 1000000000);
RANGE_CHECK(cfg, g_timebase.num, 1, cfg->g_timebase.den);
RANGE_CHECK(cfg, g_timebase.num, 1, 1000000000);
RANGE_CHECK_HI(cfg, g_profile, 3);
RANGE_CHECK_HI(cfg, rc_max_quantizer, 63);
RANGE_CHECK_HI(cfg, rc_min_quantizer, cfg->rc_max_quantizer);
@ -197,7 +199,7 @@ static vpx_codec_err_t validate_config(vpx_codec_alg_priv_t *ctx,
RANGE_CHECK_HI(vp8_cfg, arnr_strength, 6);
RANGE_CHECK(vp8_cfg, arnr_type, 1, 3);
RANGE_CHECK(vp8_cfg, cq_level, 0, 63);
RANGE_CHECK_BOOL(vp8_cfg, screen_content_mode);
RANGE_CHECK_HI(vp8_cfg, screen_content_mode, 2);
if (finalize && (cfg->rc_end_usage == VPX_CQ || cfg->rc_end_usage == VPX_Q))
RANGE_CHECK(vp8_cfg, cq_level,
cfg->rc_min_quantizer, cfg->rc_max_quantizer);
@ -365,9 +367,9 @@ static vpx_codec_err_t set_vp8e_config(VP8_CONFIG *oxcf,
if (oxcf->number_of_layers > 1)
{
memcpy (oxcf->target_bitrate, cfg.ts_target_bitrate,
sizeof(cfg.ts_target_bitrate));
sizeof(cfg.ts_target_bitrate));
memcpy (oxcf->rate_decimator, cfg.ts_rate_decimator,
sizeof(cfg.ts_rate_decimator));
sizeof(cfg.ts_rate_decimator));
memcpy (oxcf->layer_id, cfg.ts_layer_id, sizeof(cfg.ts_layer_id));
}
@ -476,8 +478,6 @@ static vpx_codec_err_t vp8e_set_config(vpx_codec_alg_priv_t *ctx,
return res;
}
int vp8_reverse_trans(int);
static vpx_codec_err_t get_quantizer(vpx_codec_alg_priv_t *ctx, va_list args)
{
int *const arg = va_arg(args, int *);
@ -649,6 +649,8 @@ static vpx_codec_err_t vp8e_init(vpx_codec_ctx_t *ctx,
vp8_rtcd();
vpx_dsp_rtcd();
vpx_scale_rtcd();
if (!ctx->priv)
{
@ -858,9 +860,6 @@ static vpx_codec_err_t vp8e_encode(vpx_codec_alg_priv_t *ctx,
{
vpx_codec_err_t res = VPX_CODEC_OK;
if (!ctx->cfg.rc_target_bitrate)
return res;
if (!ctx->cfg.rc_target_bitrate)
return res;

10
media/libvpx/vp8/vp8_dx_iface.c
Просмотреть файл

@ -11,7 +11,9 @@
#include <stdlib.h>
#include <string.h>
#include "vp8_rtcd.h"
#include "./vp8_rtcd.h"
#include "./vpx_dsp_rtcd.h"
#include "./vpx_scale_rtcd.h"
#include "vpx/vpx_decoder.h"
#include "vpx/vp8dx.h"
#include "vpx/internal/vpx_codec_internal.h"
@ -106,6 +108,8 @@ static vpx_codec_err_t vp8_init(vpx_codec_ctx_t *ctx,
(void) data;
vp8_rtcd();
vpx_dsp_rtcd();
vpx_scale_rtcd();
/* This function only allocates space for the vpx_codec_alg_priv_t
* structure. More memory may be required at the time the stream
@ -286,8 +290,8 @@ update_fragments(vpx_codec_alg_priv_t *ctx,
if (ctx->fragments.count == 0)
{
/* New frame, reset fragment pointers and sizes */
vpx_memset((void*)ctx->fragments.ptrs, 0, sizeof(ctx->fragments.ptrs));
vpx_memset(ctx->fragments.sizes, 0, sizeof(ctx->fragments.sizes));
memset((void*)ctx->fragments.ptrs, 0, sizeof(ctx->fragments.ptrs));
memset(ctx->fragments.sizes, 0, sizeof(ctx->fragments.sizes));
}
if (ctx->fragments.enabled && !(data == NULL && data_sz == 0))
{

134
media/libvpx/vp8_rtcd_armv7-android-gcc.h
Просмотреть файл

@ -33,8 +33,7 @@ RTCD_EXTERN void (*vp8_bilinear_predict16x16)(unsigned char *src, int src_pitch,
void vp8_bilinear_predict4x4_c(unsigned char *src, int src_pitch, int xofst, int yofst, unsigned char *dst, int dst_pitch);
void vp8_bilinear_predict4x4_armv6(unsigned char *src, int src_pitch, int xofst, int yofst, unsigned char *dst, int dst_pitch);
void vp8_bilinear_predict4x4_neon(unsigned char *src, int src_pitch, int xofst, int yofst, unsigned char *dst, int dst_pitch);
RTCD_EXTERN void (*vp8_bilinear_predict4x4)(unsigned char *src, int src_pitch, int xofst, int yofst, unsigned char *dst, int dst_pitch);
#define vp8_bilinear_predict4x4 vp8_bilinear_predict4x4_armv6
void vp8_bilinear_predict8x4_c(unsigned char *src, int src_pitch, int xofst, int yofst, unsigned char *dst, int dst_pitch);
void vp8_bilinear_predict8x4_armv6(unsigned char *src, int src_pitch, int xofst, int yofst, unsigned char *dst, int dst_pitch);
@ -118,13 +117,6 @@ RTCD_EXTERN void (*vp8_fast_quantize_b)(struct block *, struct blockd *);
int vp8_full_search_sad_c(struct macroblock *x, struct block *b, struct blockd *d, union int_mv *ref_mv, int sad_per_bit, int distance, struct variance_vtable *fn_ptr, int *mvcost[2], union int_mv *center_mv);
#define vp8_full_search_sad vp8_full_search_sad_c
unsigned int vp8_get4x4sse_cs_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride);
unsigned int vp8_get4x4sse_cs_neon(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride);
RTCD_EXTERN unsigned int (*vp8_get4x4sse_cs)(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride);
unsigned int vp8_get_mb_ss_c(const short *);
#define vp8_get_mb_ss vp8_get_mb_ss_c
void vp8_intra4x4_predict_c(unsigned char *Above, unsigned char *yleft, int left_stride, int b_mode, unsigned char *dst, int dst_stride, unsigned char top_left);
void vp8_intra4x4_predict_armv6(unsigned char *Above, unsigned char *yleft, int left_stride, int b_mode, unsigned char *dst, int dst_stride, unsigned char top_left);
#define vp8_intra4x4_predict vp8_intra4x4_predict_armv6
@ -175,83 +167,12 @@ int vp8_mbblock_error_c(struct macroblock *mb, int dc);
int vp8_mbuverror_c(struct macroblock *mb);
#define vp8_mbuverror vp8_mbuverror_c
unsigned int vp8_mse16x16_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_mse16x16_armv6(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_mse16x16_neon(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
RTCD_EXTERN unsigned int (*vp8_mse16x16)(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
int vp8_refining_search_sad_c(struct macroblock *x, struct block *b, struct blockd *d, union int_mv *ref_mv, int sad_per_bit, int distance, struct variance_vtable *fn_ptr, int *mvcost[2], union int_mv *center_mv);
#define vp8_refining_search_sad vp8_refining_search_sad_c
void vp8_regular_quantize_b_c(struct block *, struct blockd *);
#define vp8_regular_quantize_b vp8_regular_quantize_b_c
unsigned int vp8_sad16x16_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad16x16_armv6(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad16x16_neon(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
RTCD_EXTERN unsigned int (*vp8_sad16x16)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
void vp8_sad16x16x3_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
#define vp8_sad16x16x3 vp8_sad16x16x3_c
void vp8_sad16x16x4d_c(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
#define vp8_sad16x16x4d vp8_sad16x16x4d_c
void vp8_sad16x16x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
#define vp8_sad16x16x8 vp8_sad16x16x8_c
unsigned int vp8_sad16x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad16x8_neon(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
RTCD_EXTERN unsigned int (*vp8_sad16x8)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
void vp8_sad16x8x3_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
#define vp8_sad16x8x3 vp8_sad16x8x3_c
void vp8_sad16x8x4d_c(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
#define vp8_sad16x8x4d vp8_sad16x8x4d_c
void vp8_sad16x8x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
#define vp8_sad16x8x8 vp8_sad16x8x8_c
unsigned int vp8_sad4x4_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad4x4_neon(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
RTCD_EXTERN unsigned int (*vp8_sad4x4)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
void vp8_sad4x4x3_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
#define vp8_sad4x4x3 vp8_sad4x4x3_c
void vp8_sad4x4x4d_c(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
#define vp8_sad4x4x4d vp8_sad4x4x4d_c
void vp8_sad4x4x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
#define vp8_sad4x4x8 vp8_sad4x4x8_c
unsigned int vp8_sad8x16_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad8x16_neon(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
RTCD_EXTERN unsigned int (*vp8_sad8x16)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
void vp8_sad8x16x3_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
#define vp8_sad8x16x3 vp8_sad8x16x3_c
void vp8_sad8x16x4d_c(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
#define vp8_sad8x16x4d vp8_sad8x16x4d_c
void vp8_sad8x16x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
#define vp8_sad8x16x8 vp8_sad8x16x8_c
unsigned int vp8_sad8x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad8x8_neon(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
RTCD_EXTERN unsigned int (*vp8_sad8x8)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
void vp8_sad8x8x3_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
#define vp8_sad8x8x3 vp8_sad8x8x3_c
void vp8_sad8x8x4d_c(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
#define vp8_sad8x8x4d vp8_sad8x8x4d_c
void vp8_sad8x8x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
#define vp8_sad8x8x8 vp8_sad8x8x8_c
void vp8_short_fdct4x4_c(short *input, short *output, int pitch);
void vp8_short_fdct4x4_armv6(short *input, short *output, int pitch);
void vp8_short_fdct4x4_neon(short *input, short *output, int pitch);
@ -299,9 +220,6 @@ void vp8_sixtap_predict8x8_armv6(unsigned char *src, int src_pitch, int xofst, i
void vp8_sixtap_predict8x8_neon(unsigned char *src, int src_pitch, int xofst, int yofst, unsigned char *dst, int dst_pitch);
RTCD_EXTERN void (*vp8_sixtap_predict8x8)(unsigned char *src, int src_pitch, int xofst, int yofst, unsigned char *dst, int dst_pitch);
unsigned int vp8_sub_pixel_mse16x16_c(const unsigned char *src_ptr, int source_stride, int xoffset, int yoffset, const unsigned char *ref_ptr, int Refstride, unsigned int *sse);
#define vp8_sub_pixel_mse16x16 vp8_sub_pixel_mse16x16_c
unsigned int vp8_sub_pixel_variance16x16_c(const unsigned char *src_ptr, int source_stride, int xoffset, int yoffset, const unsigned char *ref_ptr, int Refstride, unsigned int *sse);
unsigned int vp8_sub_pixel_variance16x16_armv6(const unsigned char *src_ptr, int source_stride, int xoffset, int yoffset, const unsigned char *ref_ptr, int Refstride, unsigned int *sse);
unsigned int vp8_sub_pixel_variance16x16_neon(const unsigned char *src_ptr, int source_stride, int xoffset, int yoffset, const unsigned char *ref_ptr, int Refstride, unsigned int *sse);
@ -318,8 +236,7 @@ unsigned int vp8_sub_pixel_variance8x16_c(const unsigned char *src_ptr, int so
unsigned int vp8_sub_pixel_variance8x8_c(const unsigned char *src_ptr, int source_stride, int xoffset, int yoffset, const unsigned char *ref_ptr, int Refstride, unsigned int *sse);
unsigned int vp8_sub_pixel_variance8x8_armv6(const unsigned char *src_ptr, int source_stride, int xoffset, int yoffset, const unsigned char *ref_ptr, int Refstride, unsigned int *sse);
unsigned int vp8_sub_pixel_variance8x8_neon(const unsigned char *src_ptr, int source_stride, int xoffset, int yoffset, const unsigned char *ref_ptr, int Refstride, unsigned int *sse);
RTCD_EXTERN unsigned int (*vp8_sub_pixel_variance8x8)(const unsigned char *src_ptr, int source_stride, int xoffset, int yoffset, const unsigned char *ref_ptr, int Refstride, unsigned int *sse);
#define vp8_sub_pixel_variance8x8 vp8_sub_pixel_variance8x8_armv6
void vp8_subtract_b_c(struct block *be, struct blockd *bd, int pitch);
void vp8_subtract_b_neon(struct block *be, struct blockd *bd, int pitch);
@ -333,27 +250,6 @@ void vp8_subtract_mby_c(short *diff, unsigned char *src, int src_stride, unsigne
void vp8_subtract_mby_neon(short *diff, unsigned char *src, int src_stride, unsigned char *pred, int pred_stride);
RTCD_EXTERN void (*vp8_subtract_mby)(short *diff, unsigned char *src, int src_stride, unsigned char *pred, int pred_stride);
unsigned int vp8_variance16x16_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance16x16_armv6(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance16x16_neon(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
RTCD_EXTERN unsigned int (*vp8_variance16x16)(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance16x8_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance16x8_neon(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
RTCD_EXTERN unsigned int (*vp8_variance16x8)(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance4x4_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
#define vp8_variance4x4 vp8_variance4x4_c
unsigned int vp8_variance8x16_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance8x16_neon(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
RTCD_EXTERN unsigned int (*vp8_variance8x16)(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance8x8_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance8x8_armv6(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance8x8_neon(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
RTCD_EXTERN unsigned int (*vp8_variance8x8)(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance_halfpixvar16x16_h_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance_halfpixvar16x16_h_armv6(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance_halfpixvar16x16_h_neon(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
@ -383,8 +279,6 @@ static void setup_rtcd_internal(void)
vp8_bilinear_predict16x16 = vp8_bilinear_predict16x16_armv6;
if (flags & HAS_NEON) vp8_bilinear_predict16x16 = vp8_bilinear_predict16x16_neon;
vp8_bilinear_predict4x4 = vp8_bilinear_predict4x4_armv6;
if (flags & HAS_NEON) vp8_bilinear_predict4x4 = vp8_bilinear_predict4x4_neon;
vp8_bilinear_predict8x4 = vp8_bilinear_predict8x4_armv6;
if (flags & HAS_NEON) vp8_bilinear_predict8x4 = vp8_bilinear_predict8x4_neon;
vp8_bilinear_predict8x8 = vp8_bilinear_predict8x8_armv6;
@ -415,8 +309,6 @@ static void setup_rtcd_internal(void)
if (flags & HAS_NEON) vp8_dequantize_b = vp8_dequantize_b_neon;
vp8_fast_quantize_b = vp8_fast_quantize_b_c;
if (flags & HAS_NEON) vp8_fast_quantize_b = vp8_fast_quantize_b_neon;
vp8_get4x4sse_cs = vp8_get4x4sse_cs_c;
if (flags & HAS_NEON) vp8_get4x4sse_cs = vp8_get4x4sse_cs_neon;
vp8_loop_filter_bh = vp8_loop_filter_bh_armv6;
if (flags & HAS_NEON) vp8_loop_filter_bh = vp8_loop_filter_bh_neon;
vp8_loop_filter_bv = vp8_loop_filter_bv_armv6;
@ -433,18 +325,6 @@ static void setup_rtcd_internal(void)
if (flags & HAS_NEON) vp8_loop_filter_simple_mbh = vp8_loop_filter_mbhs_neon;
vp8_loop_filter_simple_mbv = vp8_loop_filter_simple_vertical_edge_armv6;
if (flags & HAS_NEON) vp8_loop_filter_simple_mbv = vp8_loop_filter_mbvs_neon;
vp8_mse16x16 = vp8_mse16x16_armv6;
if (flags & HAS_NEON) vp8_mse16x16 = vp8_mse16x16_neon;
vp8_sad16x16 = vp8_sad16x16_armv6;
if (flags & HAS_NEON) vp8_sad16x16 = vp8_sad16x16_neon;
vp8_sad16x8 = vp8_sad16x8_c;
if (flags & HAS_NEON) vp8_sad16x8 = vp8_sad16x8_neon;
vp8_sad4x4 = vp8_sad4x4_c;
if (flags & HAS_NEON) vp8_sad4x4 = vp8_sad4x4_neon;
vp8_sad8x16 = vp8_sad8x16_c;
if (flags & HAS_NEON) vp8_sad8x16 = vp8_sad8x16_neon;
vp8_sad8x8 = vp8_sad8x8_c;
if (flags & HAS_NEON) vp8_sad8x8 = vp8_sad8x8_neon;
vp8_short_fdct4x4 = vp8_short_fdct4x4_armv6;
if (flags & HAS_NEON) vp8_short_fdct4x4 = vp8_short_fdct4x4_neon;
vp8_short_fdct8x4 = vp8_short_fdct8x4_armv6;
@ -463,22 +343,12 @@ static void setup_rtcd_internal(void)
if (flags & HAS_NEON) vp8_sixtap_predict8x8 = vp8_sixtap_predict8x8_neon;
vp8_sub_pixel_variance16x16 = vp8_sub_pixel_variance16x16_armv6;
if (flags & HAS_NEON) vp8_sub_pixel_variance16x16 = vp8_sub_pixel_variance16x16_neon;
vp8_sub_pixel_variance8x8 = vp8_sub_pixel_variance8x8_armv6;
if (flags & HAS_NEON) vp8_sub_pixel_variance8x8 = vp8_sub_pixel_variance8x8_neon;
vp8_subtract_b = vp8_subtract_b_c;
if (flags & HAS_NEON) vp8_subtract_b = vp8_subtract_b_neon;
vp8_subtract_mbuv = vp8_subtract_mbuv_c;
if (flags & HAS_NEON) vp8_subtract_mbuv = vp8_subtract_mbuv_neon;
vp8_subtract_mby = vp8_subtract_mby_c;
if (flags & HAS_NEON) vp8_subtract_mby = vp8_subtract_mby_neon;
vp8_variance16x16 = vp8_variance16x16_armv6;
if (flags & HAS_NEON) vp8_variance16x16 = vp8_variance16x16_neon;
vp8_variance16x8 = vp8_variance16x8_c;
if (flags & HAS_NEON) vp8_variance16x8 = vp8_variance16x8_neon;
vp8_variance8x16 = vp8_variance8x16_c;
if (flags & HAS_NEON) vp8_variance8x16 = vp8_variance8x16_neon;
vp8_variance8x8 = vp8_variance8x8_armv6;
if (flags & HAS_NEON) vp8_variance8x8 = vp8_variance8x8_neon;
vp8_variance_halfpixvar16x16_h = vp8_variance_halfpixvar16x16_h_armv6;
if (flags & HAS_NEON) vp8_variance_halfpixvar16x16_h = vp8_variance_halfpixvar16x16_h_neon;
vp8_variance_halfpixvar16x16_hv = vp8_variance_halfpixvar16x16_hv_armv6;

87
media/libvpx/vp8_rtcd_generic-gnu.h
Просмотреть файл

@ -89,12 +89,6 @@ void vp8_fast_quantize_b_c(struct block *, struct blockd *);
int vp8_full_search_sad_c(struct macroblock *x, struct block *b, struct blockd *d, union int_mv *ref_mv, int sad_per_bit, int distance, struct variance_vtable *fn_ptr, int *mvcost[2], union int_mv *center_mv);
#define vp8_full_search_sad vp8_full_search_sad_c
unsigned int vp8_get4x4sse_cs_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride);
#define vp8_get4x4sse_cs vp8_get4x4sse_cs_c
unsigned int vp8_get_mb_ss_c(const short *);
#define vp8_get_mb_ss vp8_get_mb_ss_c
void vp8_intra4x4_predict_c(unsigned char *Above, unsigned char *yleft, int left_stride, int b_mode, unsigned char *dst, int dst_stride, unsigned char top_left);
#define vp8_intra4x4_predict vp8_intra4x4_predict_c
@ -128,75 +122,12 @@ int vp8_mbblock_error_c(struct macroblock *mb, int dc);
int vp8_mbuverror_c(struct macroblock *mb);
#define vp8_mbuverror vp8_mbuverror_c
unsigned int vp8_mse16x16_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
#define vp8_mse16x16 vp8_mse16x16_c
int vp8_refining_search_sad_c(struct macroblock *x, struct block *b, struct blockd *d, union int_mv *ref_mv, int sad_per_bit, int distance, struct variance_vtable *fn_ptr, int *mvcost[2], union int_mv *center_mv);
#define vp8_refining_search_sad vp8_refining_search_sad_c
void vp8_regular_quantize_b_c(struct block *, struct blockd *);
#define vp8_regular_quantize_b vp8_regular_quantize_b_c
unsigned int vp8_sad16x16_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
#define vp8_sad16x16 vp8_sad16x16_c
void vp8_sad16x16x3_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
#define vp8_sad16x16x3 vp8_sad16x16x3_c
void vp8_sad16x16x4d_c(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
#define vp8_sad16x16x4d vp8_sad16x16x4d_c
void vp8_sad16x16x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
#define vp8_sad16x16x8 vp8_sad16x16x8_c
unsigned int vp8_sad16x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
#define vp8_sad16x8 vp8_sad16x8_c
void vp8_sad16x8x3_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
#define vp8_sad16x8x3 vp8_sad16x8x3_c
void vp8_sad16x8x4d_c(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
#define vp8_sad16x8x4d vp8_sad16x8x4d_c
void vp8_sad16x8x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
#define vp8_sad16x8x8 vp8_sad16x8x8_c
unsigned int vp8_sad4x4_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
#define vp8_sad4x4 vp8_sad4x4_c
void vp8_sad4x4x3_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
#define vp8_sad4x4x3 vp8_sad4x4x3_c
void vp8_sad4x4x4d_c(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
#define vp8_sad4x4x4d vp8_sad4x4x4d_c
void vp8_sad4x4x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
#define vp8_sad4x4x8 vp8_sad4x4x8_c
unsigned int vp8_sad8x16_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
#define vp8_sad8x16 vp8_sad8x16_c
void vp8_sad8x16x3_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
#define vp8_sad8x16x3 vp8_sad8x16x3_c
void vp8_sad8x16x4d_c(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
#define vp8_sad8x16x4d vp8_sad8x16x4d_c
void vp8_sad8x16x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
#define vp8_sad8x16x8 vp8_sad8x16x8_c
unsigned int vp8_sad8x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
#define vp8_sad8x8 vp8_sad8x8_c
void vp8_sad8x8x3_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
#define vp8_sad8x8x3 vp8_sad8x8x3_c
void vp8_sad8x8x4d_c(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
#define vp8_sad8x8x4d vp8_sad8x8x4d_c
void vp8_sad8x8x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
#define vp8_sad8x8x8 vp8_sad8x8x8_c
void vp8_short_fdct4x4_c(short *input, short *output, int pitch);
#define vp8_short_fdct4x4 vp8_short_fdct4x4_c
@ -227,9 +158,6 @@ void vp8_sixtap_predict8x4_c(unsigned char *src, int src_pitch, int xofst, int y
void vp8_sixtap_predict8x8_c(unsigned char *src, int src_pitch, int xofst, int yofst, unsigned char *dst, int dst_pitch);
#define vp8_sixtap_predict8x8 vp8_sixtap_predict8x8_c
unsigned int vp8_sub_pixel_mse16x16_c(const unsigned char *src_ptr, int source_stride, int xoffset, int yoffset, const unsigned char *ref_ptr, int Refstride, unsigned int *sse);
#define vp8_sub_pixel_mse16x16 vp8_sub_pixel_mse16x16_c
unsigned int vp8_sub_pixel_variance16x16_c(const unsigned char *src_ptr, int source_stride, int xoffset, int yoffset, const unsigned char *ref_ptr, int Refstride, unsigned int *sse);
#define vp8_sub_pixel_variance16x16 vp8_sub_pixel_variance16x16_c
@ -257,21 +185,6 @@ void vp8_subtract_mby_c(short *diff, unsigned char *src, int src_stride, unsigne
void vp8_temporal_filter_apply_c(unsigned char *frame1, unsigned int stride, unsigned char *frame2, unsigned int block_size, int strength, int filter_weight, unsigned int *accumulator, unsigned short *count);
#define vp8_temporal_filter_apply vp8_temporal_filter_apply_c
unsigned int vp8_variance16x16_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
#define vp8_variance16x16 vp8_variance16x16_c
unsigned int vp8_variance16x8_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
#define vp8_variance16x8 vp8_variance16x8_c
unsigned int vp8_variance4x4_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
#define vp8_variance4x4 vp8_variance4x4_c
unsigned int vp8_variance8x16_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
#define vp8_variance8x16 vp8_variance8x16_c
unsigned int vp8_variance8x8_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
#define vp8_variance8x8 vp8_variance8x8_c
unsigned int vp8_variance_halfpixvar16x16_h_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
#define vp8_variance_halfpixvar16x16_h vp8_variance_halfpixvar16x16_h_c

214
media/libvpx/vp8_rtcd_x86-darwin9-gcc.h
Просмотреть файл

@ -74,10 +74,10 @@ void vp8_clear_system_state_c();
void vpx_reset_mmx_state();
RTCD_EXTERN void (*vp8_clear_system_state)();
void vp8_copy32xn_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, int n);
void vp8_copy32xn_sse2(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, int n);
void vp8_copy32xn_sse3(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, int n);
RTCD_EXTERN void (*vp8_copy32xn)(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, int n);
void vp8_copy32xn_c(const unsigned char *src_ptr, int source_stride, unsigned char *dst_ptr, int dst_stride, int n);
void vp8_copy32xn_sse2(const unsigned char *src_ptr, int source_stride, unsigned char *dst_ptr, int dst_stride, int n);
void vp8_copy32xn_sse3(const unsigned char *src_ptr, int source_stride, unsigned char *dst_ptr, int dst_stride, int n);
RTCD_EXTERN void (*vp8_copy32xn)(const unsigned char *src_ptr, int source_stride, unsigned char *dst_ptr, int dst_stride, int n);
void vp8_copy_mem16x16_c(unsigned char *src, int src_pitch, unsigned char *dst, int dst_pitch);
void vp8_copy_mem16x16_mmx(unsigned char *src, int src_pitch, unsigned char *dst, int dst_pitch);
@ -147,15 +147,6 @@ int vp8_full_search_sadx3(struct macroblock *x, struct block *b, struct blockd *
int vp8_full_search_sadx8(struct macroblock *x, struct block *b, struct blockd *d, union int_mv *ref_mv, int sad_per_bit, int distance, struct variance_vtable *fn_ptr, int *mvcost[2], union int_mv *center_mv);
RTCD_EXTERN int (*vp8_full_search_sad)(struct macroblock *x, struct block *b, struct blockd *d, union int_mv *ref_mv, int sad_per_bit, int distance, struct variance_vtable *fn_ptr, int *mvcost[2], union int_mv *center_mv);
unsigned int vp8_get4x4sse_cs_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride);
unsigned int vp8_get4x4sse_cs_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride);
RTCD_EXTERN unsigned int (*vp8_get4x4sse_cs)(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride);
unsigned int vp8_get_mb_ss_c(const short *);
unsigned int vp8_get_mb_ss_mmx(const short *);
unsigned int vp8_get_mb_ss_sse2(const short *);
RTCD_EXTERN unsigned int (*vp8_get_mb_ss)(const short *);
void vp8_intra4x4_predict_c(unsigned char *Above, unsigned char *yleft, int left_stride, int b_mode, unsigned char *dst, int dst_stride, unsigned char top_left);
#define vp8_intra4x4_predict vp8_intra4x4_predict_c
@ -218,11 +209,6 @@ int vp8_mbuverror_mmx(struct macroblock *mb);
int vp8_mbuverror_xmm(struct macroblock *mb);
RTCD_EXTERN int (*vp8_mbuverror)(struct macroblock *mb);
unsigned int vp8_mse16x16_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_mse16x16_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_mse16x16_wmt(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
RTCD_EXTERN unsigned int (*vp8_mse16x16)(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
void vp8_plane_add_noise_c(unsigned char *s, char *noise, char blackclamp[16], char whiteclamp[16], char bothclamp[16], unsigned int w, unsigned int h, int pitch);
void vp8_plane_add_noise_mmx(unsigned char *s, char *noise, char blackclamp[16], char whiteclamp[16], char bothclamp[16], unsigned int w, unsigned int h, int pitch);
void vp8_plane_add_noise_wmt(unsigned char *s, char *noise, char blackclamp[16], char whiteclamp[16], char bothclamp[16], unsigned int w, unsigned int h, int pitch);
@ -241,94 +227,6 @@ void vp8_regular_quantize_b_sse2(struct block *, struct blockd *);
void vp8_regular_quantize_b_sse4_1(struct block *, struct blockd *);
RTCD_EXTERN void (*vp8_regular_quantize_b)(struct block *, struct blockd *);
unsigned int vp8_sad16x16_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad16x16_mmx(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad16x16_wmt(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad16x16_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
RTCD_EXTERN unsigned int (*vp8_sad16x16)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
void vp8_sad16x16x3_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad16x16x3_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad16x16x3_ssse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad16x16x3)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad16x16x4d_c(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad16x16x4d_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad16x16x4d)(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad16x16x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
void vp8_sad16x16x8_sse4(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
RTCD_EXTERN void (*vp8_sad16x16x8)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
unsigned int vp8_sad16x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad16x8_mmx(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad16x8_wmt(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
RTCD_EXTERN unsigned int (*vp8_sad16x8)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
void vp8_sad16x8x3_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad16x8x3_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad16x8x3_ssse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad16x8x3)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad16x8x4d_c(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad16x8x4d_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad16x8x4d)(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad16x8x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
void vp8_sad16x8x8_sse4(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
RTCD_EXTERN void (*vp8_sad16x8x8)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
unsigned int vp8_sad4x4_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad4x4_mmx(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad4x4_wmt(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
RTCD_EXTERN unsigned int (*vp8_sad4x4)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
void vp8_sad4x4x3_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad4x4x3_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad4x4x3)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad4x4x4d_c(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad4x4x4d_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad4x4x4d)(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad4x4x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
void vp8_sad4x4x8_sse4(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
RTCD_EXTERN void (*vp8_sad4x4x8)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
unsigned int vp8_sad8x16_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad8x16_mmx(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad8x16_wmt(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
RTCD_EXTERN unsigned int (*vp8_sad8x16)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
void vp8_sad8x16x3_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad8x16x3_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad8x16x3)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad8x16x4d_c(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad8x16x4d_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad8x16x4d)(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad8x16x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
void vp8_sad8x16x8_sse4(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
RTCD_EXTERN void (*vp8_sad8x16x8)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
unsigned int vp8_sad8x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad8x8_mmx(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad8x8_wmt(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
RTCD_EXTERN unsigned int (*vp8_sad8x8)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
void vp8_sad8x8x3_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad8x8x3_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad8x8x3)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad8x8x4d_c(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad8x8x4d_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad8x8x4d)(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad8x8x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
void vp8_sad8x8x8_sse4(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
RTCD_EXTERN void (*vp8_sad8x8x8)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
void vp8_short_fdct4x4_c(short *input, short *output, int pitch);
void vp8_short_fdct4x4_mmx(short *input, short *output, int pitch);
void vp8_short_fdct4x4_sse2(short *input, short *output, int pitch);
@ -378,11 +276,6 @@ void vp8_sixtap_predict8x8_sse2(unsigned char *src, int src_pitch, int xofst, in
void vp8_sixtap_predict8x8_ssse3(unsigned char *src, int src_pitch, int xofst, int yofst, unsigned char *dst, int dst_pitch);
RTCD_EXTERN void (*vp8_sixtap_predict8x8)(unsigned char *src, int src_pitch, int xofst, int yofst, unsigned char *dst, int dst_pitch);
unsigned int vp8_sub_pixel_mse16x16_c(const unsigned char *src_ptr, int source_stride, int xoffset, int yoffset, const unsigned char *ref_ptr, int Refstride, unsigned int *sse);
unsigned int vp8_sub_pixel_mse16x16_mmx(const unsigned char *src_ptr, int source_stride, int xoffset, int yoffset, const unsigned char *ref_ptr, int Refstride, unsigned int *sse);
unsigned int vp8_sub_pixel_mse16x16_wmt(const unsigned char *src_ptr, int source_stride, int xoffset, int yoffset, const unsigned char *ref_ptr, int Refstride, unsigned int *sse);
RTCD_EXTERN unsigned int (*vp8_sub_pixel_mse16x16)(const unsigned char *src_ptr, int source_stride, int xoffset, int yoffset, const unsigned char *ref_ptr, int Refstride, unsigned int *sse);
unsigned int vp8_sub_pixel_variance16x16_c(const unsigned char *src_ptr, int source_stride, int xoffset, int yoffset, const unsigned char *ref_ptr, int Refstride, unsigned int *sse);
unsigned int vp8_sub_pixel_variance16x16_mmx(const unsigned char *src_ptr, int source_stride, int xoffset, int yoffset, const unsigned char *ref_ptr, int Refstride, unsigned int *sse);
unsigned int vp8_sub_pixel_variance16x16_wmt(const unsigned char *src_ptr, int source_stride, int xoffset, int yoffset, const unsigned char *ref_ptr, int Refstride, unsigned int *sse);
@ -429,31 +322,6 @@ void vp8_temporal_filter_apply_c(unsigned char *frame1, unsigned int stride, uns
void vp8_temporal_filter_apply_sse2(unsigned char *frame1, unsigned int stride, unsigned char *frame2, unsigned int block_size, int strength, int filter_weight, unsigned int *accumulator, unsigned short *count);
RTCD_EXTERN void (*vp8_temporal_filter_apply)(unsigned char *frame1, unsigned int stride, unsigned char *frame2, unsigned int block_size, int strength, int filter_weight, unsigned int *accumulator, unsigned short *count);
unsigned int vp8_variance16x16_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance16x16_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance16x16_wmt(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
RTCD_EXTERN unsigned int (*vp8_variance16x16)(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance16x8_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance16x8_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance16x8_wmt(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
RTCD_EXTERN unsigned int (*vp8_variance16x8)(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance4x4_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance4x4_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance4x4_wmt(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
RTCD_EXTERN unsigned int (*vp8_variance4x4)(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance8x16_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance8x16_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance8x16_wmt(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
RTCD_EXTERN unsigned int (*vp8_variance8x16)(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance8x8_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance8x8_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance8x8_wmt(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
RTCD_EXTERN unsigned int (*vp8_variance8x8)(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance_halfpixvar16x16_h_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance_halfpixvar16x16_h_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance_halfpixvar16x16_h_wmt(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
@ -540,11 +408,6 @@ static void setup_rtcd_internal(void)
vp8_full_search_sad = vp8_full_search_sad_c;
if (flags & HAS_SSE3) vp8_full_search_sad = vp8_full_search_sadx3;
if (flags & HAS_SSE4_1) vp8_full_search_sad = vp8_full_search_sadx8;
vp8_get4x4sse_cs = vp8_get4x4sse_cs_c;
if (flags & HAS_MMX) vp8_get4x4sse_cs = vp8_get4x4sse_cs_mmx;
vp8_get_mb_ss = vp8_get_mb_ss_c;
if (flags & HAS_MMX) vp8_get_mb_ss = vp8_get_mb_ss_mmx;
if (flags & HAS_SSE2) vp8_get_mb_ss = vp8_get_mb_ss_sse2;
vp8_loop_filter_bh = vp8_loop_filter_bh_c;
if (flags & HAS_MMX) vp8_loop_filter_bh = vp8_loop_filter_bh_mmx;
if (flags & HAS_SSE2) vp8_loop_filter_bh = vp8_loop_filter_bh_sse2;
@ -580,9 +443,6 @@ static void setup_rtcd_internal(void)
vp8_mbuverror = vp8_mbuverror_c;
if (flags & HAS_MMX) vp8_mbuverror = vp8_mbuverror_mmx;
if (flags & HAS_SSE2) vp8_mbuverror = vp8_mbuverror_xmm;
vp8_mse16x16 = vp8_mse16x16_c;
if (flags & HAS_MMX) vp8_mse16x16 = vp8_mse16x16_mmx;
if (flags & HAS_SSE2) vp8_mse16x16 = vp8_mse16x16_wmt;
vp8_plane_add_noise = vp8_plane_add_noise_c;
if (flags & HAS_MMX) vp8_plane_add_noise = vp8_plane_add_noise_mmx;
if (flags & HAS_SSE2) vp8_plane_add_noise = vp8_plane_add_noise_wmt;
@ -593,54 +453,6 @@ static void setup_rtcd_internal(void)
vp8_regular_quantize_b = vp8_regular_quantize_b_c;
if (flags & HAS_SSE2) vp8_regular_quantize_b = vp8_regular_quantize_b_sse2;
if (flags & HAS_SSE4_1) vp8_regular_quantize_b = vp8_regular_quantize_b_sse4_1;
vp8_sad16x16 = vp8_sad16x16_c;
if (flags & HAS_MMX) vp8_sad16x16 = vp8_sad16x16_mmx;
if (flags & HAS_SSE2) vp8_sad16x16 = vp8_sad16x16_wmt;
if (flags & HAS_SSE3) vp8_sad16x16 = vp8_sad16x16_sse3;
vp8_sad16x16x3 = vp8_sad16x16x3_c;
if (flags & HAS_SSE3) vp8_sad16x16x3 = vp8_sad16x16x3_sse3;
if (flags & HAS_SSSE3) vp8_sad16x16x3 = vp8_sad16x16x3_ssse3;
vp8_sad16x16x4d = vp8_sad16x16x4d_c;
if (flags & HAS_SSE3) vp8_sad16x16x4d = vp8_sad16x16x4d_sse3;
vp8_sad16x16x8 = vp8_sad16x16x8_c;
if (flags & HAS_SSE4_1) vp8_sad16x16x8 = vp8_sad16x16x8_sse4;
vp8_sad16x8 = vp8_sad16x8_c;
if (flags & HAS_MMX) vp8_sad16x8 = vp8_sad16x8_mmx;
if (flags & HAS_SSE2) vp8_sad16x8 = vp8_sad16x8_wmt;
vp8_sad16x8x3 = vp8_sad16x8x3_c;
if (flags & HAS_SSE3) vp8_sad16x8x3 = vp8_sad16x8x3_sse3;
if (flags & HAS_SSSE3) vp8_sad16x8x3 = vp8_sad16x8x3_ssse3;
vp8_sad16x8x4d = vp8_sad16x8x4d_c;
if (flags & HAS_SSE3) vp8_sad16x8x4d = vp8_sad16x8x4d_sse3;
vp8_sad16x8x8 = vp8_sad16x8x8_c;
if (flags & HAS_SSE4_1) vp8_sad16x8x8 = vp8_sad16x8x8_sse4;
vp8_sad4x4 = vp8_sad4x4_c;
if (flags & HAS_MMX) vp8_sad4x4 = vp8_sad4x4_mmx;
if (flags & HAS_SSE2) vp8_sad4x4 = vp8_sad4x4_wmt;
vp8_sad4x4x3 = vp8_sad4x4x3_c;
if (flags & HAS_SSE3) vp8_sad4x4x3 = vp8_sad4x4x3_sse3;
vp8_sad4x4x4d = vp8_sad4x4x4d_c;
if (flags & HAS_SSE3) vp8_sad4x4x4d = vp8_sad4x4x4d_sse3;
vp8_sad4x4x8 = vp8_sad4x4x8_c;
if (flags & HAS_SSE4_1) vp8_sad4x4x8 = vp8_sad4x4x8_sse4;
vp8_sad8x16 = vp8_sad8x16_c;
if (flags & HAS_MMX) vp8_sad8x16 = vp8_sad8x16_mmx;
if (flags & HAS_SSE2) vp8_sad8x16 = vp8_sad8x16_wmt;
vp8_sad8x16x3 = vp8_sad8x16x3_c;
if (flags & HAS_SSE3) vp8_sad8x16x3 = vp8_sad8x16x3_sse3;
vp8_sad8x16x4d = vp8_sad8x16x4d_c;
if (flags & HAS_SSE3) vp8_sad8x16x4d = vp8_sad8x16x4d_sse3;
vp8_sad8x16x8 = vp8_sad8x16x8_c;
if (flags & HAS_SSE4_1) vp8_sad8x16x8 = vp8_sad8x16x8_sse4;
vp8_sad8x8 = vp8_sad8x8_c;
if (flags & HAS_MMX) vp8_sad8x8 = vp8_sad8x8_mmx;
if (flags & HAS_SSE2) vp8_sad8x8 = vp8_sad8x8_wmt;
vp8_sad8x8x3 = vp8_sad8x8x3_c;
if (flags & HAS_SSE3) vp8_sad8x8x3 = vp8_sad8x8x3_sse3;
vp8_sad8x8x4d = vp8_sad8x8x4d_c;
if (flags & HAS_SSE3) vp8_sad8x8x4d = vp8_sad8x8x4d_sse3;
vp8_sad8x8x8 = vp8_sad8x8x8_c;
if (flags & HAS_SSE4_1) vp8_sad8x8x8 = vp8_sad8x8x8_sse4;
vp8_short_fdct4x4 = vp8_short_fdct4x4_c;
if (flags & HAS_MMX) vp8_short_fdct4x4 = vp8_short_fdct4x4_mmx;
if (flags & HAS_SSE2) vp8_short_fdct4x4 = vp8_short_fdct4x4_sse2;
@ -669,9 +481,6 @@ static void setup_rtcd_internal(void)
if (flags & HAS_MMX) vp8_sixtap_predict8x8 = vp8_sixtap_predict8x8_mmx;
if (flags & HAS_SSE2) vp8_sixtap_predict8x8 = vp8_sixtap_predict8x8_sse2;
if (flags & HAS_SSSE3) vp8_sixtap_predict8x8 = vp8_sixtap_predict8x8_ssse3;
vp8_sub_pixel_mse16x16 = vp8_sub_pixel_mse16x16_c;
if (flags & HAS_MMX) vp8_sub_pixel_mse16x16 = vp8_sub_pixel_mse16x16_mmx;
if (flags & HAS_SSE2) vp8_sub_pixel_mse16x16 = vp8_sub_pixel_mse16x16_wmt;
vp8_sub_pixel_variance16x16 = vp8_sub_pixel_variance16x16_c;
if (flags & HAS_MMX) vp8_sub_pixel_variance16x16 = vp8_sub_pixel_variance16x16_mmx;
if (flags & HAS_SSE2) vp8_sub_pixel_variance16x16 = vp8_sub_pixel_variance16x16_wmt;
@ -700,21 +509,6 @@ static void setup_rtcd_internal(void)
if (flags & HAS_SSE2) vp8_subtract_mby = vp8_subtract_mby_sse2;
vp8_temporal_filter_apply = vp8_temporal_filter_apply_c;
if (flags & HAS_SSE2) vp8_temporal_filter_apply = vp8_temporal_filter_apply_sse2;
vp8_variance16x16 = vp8_variance16x16_c;
if (flags & HAS_MMX) vp8_variance16x16 = vp8_variance16x16_mmx;
if (flags & HAS_SSE2) vp8_variance16x16 = vp8_variance16x16_wmt;
vp8_variance16x8 = vp8_variance16x8_c;
if (flags & HAS_MMX) vp8_variance16x8 = vp8_variance16x8_mmx;
if (flags & HAS_SSE2) vp8_variance16x8 = vp8_variance16x8_wmt;
vp8_variance4x4 = vp8_variance4x4_c;
if (flags & HAS_MMX) vp8_variance4x4 = vp8_variance4x4_mmx;
if (flags & HAS_SSE2) vp8_variance4x4 = vp8_variance4x4_wmt;
vp8_variance8x16 = vp8_variance8x16_c;
if (flags & HAS_MMX) vp8_variance8x16 = vp8_variance8x16_mmx;
if (flags & HAS_SSE2) vp8_variance8x16 = vp8_variance8x16_wmt;
vp8_variance8x8 = vp8_variance8x8_c;
if (flags & HAS_MMX) vp8_variance8x8 = vp8_variance8x8_mmx;
if (flags & HAS_SSE2) vp8_variance8x8 = vp8_variance8x8_wmt;
vp8_variance_halfpixvar16x16_h = vp8_variance_halfpixvar16x16_h_c;
if (flags & HAS_MMX) vp8_variance_halfpixvar16x16_h = vp8_variance_halfpixvar16x16_h_mmx;
if (flags & HAS_SSE2) vp8_variance_halfpixvar16x16_h = vp8_variance_halfpixvar16x16_h_wmt;

214
media/libvpx/vp8_rtcd_x86-linux-gcc.h
Просмотреть файл

@ -74,10 +74,10 @@ void vp8_clear_system_state_c();
void vpx_reset_mmx_state();
RTCD_EXTERN void (*vp8_clear_system_state)();
void vp8_copy32xn_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, int n);
void vp8_copy32xn_sse2(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, int n);
void vp8_copy32xn_sse3(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, int n);
RTCD_EXTERN void (*vp8_copy32xn)(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, int n);
void vp8_copy32xn_c(const unsigned char *src_ptr, int source_stride, unsigned char *dst_ptr, int dst_stride, int n);
void vp8_copy32xn_sse2(const unsigned char *src_ptr, int source_stride, unsigned char *dst_ptr, int dst_stride, int n);
void vp8_copy32xn_sse3(const unsigned char *src_ptr, int source_stride, unsigned char *dst_ptr, int dst_stride, int n);
RTCD_EXTERN void (*vp8_copy32xn)(const unsigned char *src_ptr, int source_stride, unsigned char *dst_ptr, int dst_stride, int n);
void vp8_copy_mem16x16_c(unsigned char *src, int src_pitch, unsigned char *dst, int dst_pitch);
void vp8_copy_mem16x16_mmx(unsigned char *src, int src_pitch, unsigned char *dst, int dst_pitch);
@ -147,15 +147,6 @@ int vp8_full_search_sadx3(struct macroblock *x, struct block *b, struct blockd *
int vp8_full_search_sadx8(struct macroblock *x, struct block *b, struct blockd *d, union int_mv *ref_mv, int sad_per_bit, int distance, struct variance_vtable *fn_ptr, int *mvcost[2], union int_mv *center_mv);
RTCD_EXTERN int (*vp8_full_search_sad)(struct macroblock *x, struct block *b, struct blockd *d, union int_mv *ref_mv, int sad_per_bit, int distance, struct variance_vtable *fn_ptr, int *mvcost[2], union int_mv *center_mv);
unsigned int vp8_get4x4sse_cs_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride);
unsigned int vp8_get4x4sse_cs_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride);
RTCD_EXTERN unsigned int (*vp8_get4x4sse_cs)(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride);
unsigned int vp8_get_mb_ss_c(const short *);
unsigned int vp8_get_mb_ss_mmx(const short *);
unsigned int vp8_get_mb_ss_sse2(const short *);
RTCD_EXTERN unsigned int (*vp8_get_mb_ss)(const short *);
void vp8_intra4x4_predict_c(unsigned char *Above, unsigned char *yleft, int left_stride, int b_mode, unsigned char *dst, int dst_stride, unsigned char top_left);
#define vp8_intra4x4_predict vp8_intra4x4_predict_c
@ -218,11 +209,6 @@ int vp8_mbuverror_mmx(struct macroblock *mb);
int vp8_mbuverror_xmm(struct macroblock *mb);
RTCD_EXTERN int (*vp8_mbuverror)(struct macroblock *mb);
unsigned int vp8_mse16x16_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_mse16x16_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_mse16x16_wmt(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
RTCD_EXTERN unsigned int (*vp8_mse16x16)(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
void vp8_plane_add_noise_c(unsigned char *s, char *noise, char blackclamp[16], char whiteclamp[16], char bothclamp[16], unsigned int w, unsigned int h, int pitch);
void vp8_plane_add_noise_mmx(unsigned char *s, char *noise, char blackclamp[16], char whiteclamp[16], char bothclamp[16], unsigned int w, unsigned int h, int pitch);
void vp8_plane_add_noise_wmt(unsigned char *s, char *noise, char blackclamp[16], char whiteclamp[16], char bothclamp[16], unsigned int w, unsigned int h, int pitch);
@ -241,94 +227,6 @@ void vp8_regular_quantize_b_sse2(struct block *, struct blockd *);
void vp8_regular_quantize_b_sse4_1(struct block *, struct blockd *);
RTCD_EXTERN void (*vp8_regular_quantize_b)(struct block *, struct blockd *);
unsigned int vp8_sad16x16_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad16x16_mmx(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad16x16_wmt(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad16x16_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
RTCD_EXTERN unsigned int (*vp8_sad16x16)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
void vp8_sad16x16x3_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad16x16x3_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad16x16x3_ssse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad16x16x3)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad16x16x4d_c(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad16x16x4d_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad16x16x4d)(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad16x16x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
void vp8_sad16x16x8_sse4(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
RTCD_EXTERN void (*vp8_sad16x16x8)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
unsigned int vp8_sad16x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad16x8_mmx(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad16x8_wmt(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
RTCD_EXTERN unsigned int (*vp8_sad16x8)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
void vp8_sad16x8x3_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad16x8x3_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad16x8x3_ssse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad16x8x3)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad16x8x4d_c(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad16x8x4d_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad16x8x4d)(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad16x8x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
void vp8_sad16x8x8_sse4(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
RTCD_EXTERN void (*vp8_sad16x8x8)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
unsigned int vp8_sad4x4_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad4x4_mmx(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad4x4_wmt(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
RTCD_EXTERN unsigned int (*vp8_sad4x4)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
void vp8_sad4x4x3_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad4x4x3_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad4x4x3)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad4x4x4d_c(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad4x4x4d_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad4x4x4d)(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad4x4x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
void vp8_sad4x4x8_sse4(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
RTCD_EXTERN void (*vp8_sad4x4x8)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
unsigned int vp8_sad8x16_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad8x16_mmx(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad8x16_wmt(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
RTCD_EXTERN unsigned int (*vp8_sad8x16)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
void vp8_sad8x16x3_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad8x16x3_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad8x16x3)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad8x16x4d_c(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad8x16x4d_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad8x16x4d)(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad8x16x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
void vp8_sad8x16x8_sse4(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
RTCD_EXTERN void (*vp8_sad8x16x8)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
unsigned int vp8_sad8x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad8x8_mmx(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad8x8_wmt(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
RTCD_EXTERN unsigned int (*vp8_sad8x8)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
void vp8_sad8x8x3_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad8x8x3_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad8x8x3)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad8x8x4d_c(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad8x8x4d_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad8x8x4d)(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad8x8x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
void vp8_sad8x8x8_sse4(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
RTCD_EXTERN void (*vp8_sad8x8x8)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
void vp8_short_fdct4x4_c(short *input, short *output, int pitch);
void vp8_short_fdct4x4_mmx(short *input, short *output, int pitch);
void vp8_short_fdct4x4_sse2(short *input, short *output, int pitch);
@ -378,11 +276,6 @@ void vp8_sixtap_predict8x8_sse2(unsigned char *src, int src_pitch, int xofst, in
void vp8_sixtap_predict8x8_ssse3(unsigned char *src, int src_pitch, int xofst, int yofst, unsigned char *dst, int dst_pitch);
RTCD_EXTERN void (*vp8_sixtap_predict8x8)(unsigned char *src, int src_pitch, int xofst, int yofst, unsigned char *dst, int dst_pitch);
unsigned int vp8_sub_pixel_mse16x16_c(const unsigned char *src_ptr, int source_stride, int xoffset, int yoffset, const unsigned char *ref_ptr, int Refstride, unsigned int *sse);
unsigned int vp8_sub_pixel_mse16x16_mmx(const unsigned char *src_ptr, int source_stride, int xoffset, int yoffset, const unsigned char *ref_ptr, int Refstride, unsigned int *sse);
unsigned int vp8_sub_pixel_mse16x16_wmt(const unsigned char *src_ptr, int source_stride, int xoffset, int yoffset, const unsigned char *ref_ptr, int Refstride, unsigned int *sse);
RTCD_EXTERN unsigned int (*vp8_sub_pixel_mse16x16)(const unsigned char *src_ptr, int source_stride, int xoffset, int yoffset, const unsigned char *ref_ptr, int Refstride, unsigned int *sse);
unsigned int vp8_sub_pixel_variance16x16_c(const unsigned char *src_ptr, int source_stride, int xoffset, int yoffset, const unsigned char *ref_ptr, int Refstride, unsigned int *sse);
unsigned int vp8_sub_pixel_variance16x16_mmx(const unsigned char *src_ptr, int source_stride, int xoffset, int yoffset, const unsigned char *ref_ptr, int Refstride, unsigned int *sse);
unsigned int vp8_sub_pixel_variance16x16_wmt(const unsigned char *src_ptr, int source_stride, int xoffset, int yoffset, const unsigned char *ref_ptr, int Refstride, unsigned int *sse);
@ -429,31 +322,6 @@ void vp8_temporal_filter_apply_c(unsigned char *frame1, unsigned int stride, uns
void vp8_temporal_filter_apply_sse2(unsigned char *frame1, unsigned int stride, unsigned char *frame2, unsigned int block_size, int strength, int filter_weight, unsigned int *accumulator, unsigned short *count);
RTCD_EXTERN void (*vp8_temporal_filter_apply)(unsigned char *frame1, unsigned int stride, unsigned char *frame2, unsigned int block_size, int strength, int filter_weight, unsigned int *accumulator, unsigned short *count);
unsigned int vp8_variance16x16_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance16x16_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance16x16_wmt(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
RTCD_EXTERN unsigned int (*vp8_variance16x16)(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance16x8_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance16x8_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance16x8_wmt(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
RTCD_EXTERN unsigned int (*vp8_variance16x8)(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance4x4_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance4x4_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance4x4_wmt(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
RTCD_EXTERN unsigned int (*vp8_variance4x4)(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance8x16_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance8x16_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance8x16_wmt(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
RTCD_EXTERN unsigned int (*vp8_variance8x16)(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance8x8_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance8x8_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance8x8_wmt(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
RTCD_EXTERN unsigned int (*vp8_variance8x8)(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance_halfpixvar16x16_h_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance_halfpixvar16x16_h_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance_halfpixvar16x16_h_wmt(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
@ -540,11 +408,6 @@ static void setup_rtcd_internal(void)
vp8_full_search_sad = vp8_full_search_sad_c;
if (flags & HAS_SSE3) vp8_full_search_sad = vp8_full_search_sadx3;
if (flags & HAS_SSE4_1) vp8_full_search_sad = vp8_full_search_sadx8;
vp8_get4x4sse_cs = vp8_get4x4sse_cs_c;
if (flags & HAS_MMX) vp8_get4x4sse_cs = vp8_get4x4sse_cs_mmx;
vp8_get_mb_ss = vp8_get_mb_ss_c;
if (flags & HAS_MMX) vp8_get_mb_ss = vp8_get_mb_ss_mmx;
if (flags & HAS_SSE2) vp8_get_mb_ss = vp8_get_mb_ss_sse2;
vp8_loop_filter_bh = vp8_loop_filter_bh_c;
if (flags & HAS_MMX) vp8_loop_filter_bh = vp8_loop_filter_bh_mmx;
if (flags & HAS_SSE2) vp8_loop_filter_bh = vp8_loop_filter_bh_sse2;
@ -580,9 +443,6 @@ static void setup_rtcd_internal(void)
vp8_mbuverror = vp8_mbuverror_c;
if (flags & HAS_MMX) vp8_mbuverror = vp8_mbuverror_mmx;
if (flags & HAS_SSE2) vp8_mbuverror = vp8_mbuverror_xmm;
vp8_mse16x16 = vp8_mse16x16_c;
if (flags & HAS_MMX) vp8_mse16x16 = vp8_mse16x16_mmx;
if (flags & HAS_SSE2) vp8_mse16x16 = vp8_mse16x16_wmt;
vp8_plane_add_noise = vp8_plane_add_noise_c;
if (flags & HAS_MMX) vp8_plane_add_noise = vp8_plane_add_noise_mmx;
if (flags & HAS_SSE2) vp8_plane_add_noise = vp8_plane_add_noise_wmt;
@ -593,54 +453,6 @@ static void setup_rtcd_internal(void)
vp8_regular_quantize_b = vp8_regular_quantize_b_c;
if (flags & HAS_SSE2) vp8_regular_quantize_b = vp8_regular_quantize_b_sse2;
if (flags & HAS_SSE4_1) vp8_regular_quantize_b = vp8_regular_quantize_b_sse4_1;
vp8_sad16x16 = vp8_sad16x16_c;
if (flags & HAS_MMX) vp8_sad16x16 = vp8_sad16x16_mmx;
if (flags & HAS_SSE2) vp8_sad16x16 = vp8_sad16x16_wmt;
if (flags & HAS_SSE3) vp8_sad16x16 = vp8_sad16x16_sse3;
vp8_sad16x16x3 = vp8_sad16x16x3_c;
if (flags & HAS_SSE3) vp8_sad16x16x3 = vp8_sad16x16x3_sse3;
if (flags & HAS_SSSE3) vp8_sad16x16x3 = vp8_sad16x16x3_ssse3;
vp8_sad16x16x4d = vp8_sad16x16x4d_c;
if (flags & HAS_SSE3) vp8_sad16x16x4d = vp8_sad16x16x4d_sse3;
vp8_sad16x16x8 = vp8_sad16x16x8_c;
if (flags & HAS_SSE4_1) vp8_sad16x16x8 = vp8_sad16x16x8_sse4;
vp8_sad16x8 = vp8_sad16x8_c;
if (flags & HAS_MMX) vp8_sad16x8 = vp8_sad16x8_mmx;
if (flags & HAS_SSE2) vp8_sad16x8 = vp8_sad16x8_wmt;
vp8_sad16x8x3 = vp8_sad16x8x3_c;
if (flags & HAS_SSE3) vp8_sad16x8x3 = vp8_sad16x8x3_sse3;
if (flags & HAS_SSSE3) vp8_sad16x8x3 = vp8_sad16x8x3_ssse3;
vp8_sad16x8x4d = vp8_sad16x8x4d_c;
if (flags & HAS_SSE3) vp8_sad16x8x4d = vp8_sad16x8x4d_sse3;
vp8_sad16x8x8 = vp8_sad16x8x8_c;
if (flags & HAS_SSE4_1) vp8_sad16x8x8 = vp8_sad16x8x8_sse4;
vp8_sad4x4 = vp8_sad4x4_c;
if (flags & HAS_MMX) vp8_sad4x4 = vp8_sad4x4_mmx;
if (flags & HAS_SSE2) vp8_sad4x4 = vp8_sad4x4_wmt;
vp8_sad4x4x3 = vp8_sad4x4x3_c;
if (flags & HAS_SSE3) vp8_sad4x4x3 = vp8_sad4x4x3_sse3;
vp8_sad4x4x4d = vp8_sad4x4x4d_c;
if (flags & HAS_SSE3) vp8_sad4x4x4d = vp8_sad4x4x4d_sse3;
vp8_sad4x4x8 = vp8_sad4x4x8_c;
if (flags & HAS_SSE4_1) vp8_sad4x4x8 = vp8_sad4x4x8_sse4;
vp8_sad8x16 = vp8_sad8x16_c;
if (flags & HAS_MMX) vp8_sad8x16 = vp8_sad8x16_mmx;
if (flags & HAS_SSE2) vp8_sad8x16 = vp8_sad8x16_wmt;
vp8_sad8x16x3 = vp8_sad8x16x3_c;
if (flags & HAS_SSE3) vp8_sad8x16x3 = vp8_sad8x16x3_sse3;
vp8_sad8x16x4d = vp8_sad8x16x4d_c;
if (flags & HAS_SSE3) vp8_sad8x16x4d = vp8_sad8x16x4d_sse3;
vp8_sad8x16x8 = vp8_sad8x16x8_c;
if (flags & HAS_SSE4_1) vp8_sad8x16x8 = vp8_sad8x16x8_sse4;
vp8_sad8x8 = vp8_sad8x8_c;
if (flags & HAS_MMX) vp8_sad8x8 = vp8_sad8x8_mmx;
if (flags & HAS_SSE2) vp8_sad8x8 = vp8_sad8x8_wmt;
vp8_sad8x8x3 = vp8_sad8x8x3_c;
if (flags & HAS_SSE3) vp8_sad8x8x3 = vp8_sad8x8x3_sse3;
vp8_sad8x8x4d = vp8_sad8x8x4d_c;
if (flags & HAS_SSE3) vp8_sad8x8x4d = vp8_sad8x8x4d_sse3;
vp8_sad8x8x8 = vp8_sad8x8x8_c;
if (flags & HAS_SSE4_1) vp8_sad8x8x8 = vp8_sad8x8x8_sse4;
vp8_short_fdct4x4 = vp8_short_fdct4x4_c;
if (flags & HAS_MMX) vp8_short_fdct4x4 = vp8_short_fdct4x4_mmx;
if (flags & HAS_SSE2) vp8_short_fdct4x4 = vp8_short_fdct4x4_sse2;
@ -669,9 +481,6 @@ static void setup_rtcd_internal(void)
if (flags & HAS_MMX) vp8_sixtap_predict8x8 = vp8_sixtap_predict8x8_mmx;
if (flags & HAS_SSE2) vp8_sixtap_predict8x8 = vp8_sixtap_predict8x8_sse2;
if (flags & HAS_SSSE3) vp8_sixtap_predict8x8 = vp8_sixtap_predict8x8_ssse3;
vp8_sub_pixel_mse16x16 = vp8_sub_pixel_mse16x16_c;
if (flags & HAS_MMX) vp8_sub_pixel_mse16x16 = vp8_sub_pixel_mse16x16_mmx;
if (flags & HAS_SSE2) vp8_sub_pixel_mse16x16 = vp8_sub_pixel_mse16x16_wmt;
vp8_sub_pixel_variance16x16 = vp8_sub_pixel_variance16x16_c;
if (flags & HAS_MMX) vp8_sub_pixel_variance16x16 = vp8_sub_pixel_variance16x16_mmx;
if (flags & HAS_SSE2) vp8_sub_pixel_variance16x16 = vp8_sub_pixel_variance16x16_wmt;
@ -700,21 +509,6 @@ static void setup_rtcd_internal(void)
if (flags & HAS_SSE2) vp8_subtract_mby = vp8_subtract_mby_sse2;
vp8_temporal_filter_apply = vp8_temporal_filter_apply_c;
if (flags & HAS_SSE2) vp8_temporal_filter_apply = vp8_temporal_filter_apply_sse2;
vp8_variance16x16 = vp8_variance16x16_c;
if (flags & HAS_MMX) vp8_variance16x16 = vp8_variance16x16_mmx;
if (flags & HAS_SSE2) vp8_variance16x16 = vp8_variance16x16_wmt;
vp8_variance16x8 = vp8_variance16x8_c;
if (flags & HAS_MMX) vp8_variance16x8 = vp8_variance16x8_mmx;
if (flags & HAS_SSE2) vp8_variance16x8 = vp8_variance16x8_wmt;
vp8_variance4x4 = vp8_variance4x4_c;
if (flags & HAS_MMX) vp8_variance4x4 = vp8_variance4x4_mmx;
if (flags & HAS_SSE2) vp8_variance4x4 = vp8_variance4x4_wmt;
vp8_variance8x16 = vp8_variance8x16_c;
if (flags & HAS_MMX) vp8_variance8x16 = vp8_variance8x16_mmx;
if (flags & HAS_SSE2) vp8_variance8x16 = vp8_variance8x16_wmt;
vp8_variance8x8 = vp8_variance8x8_c;
if (flags & HAS_MMX) vp8_variance8x8 = vp8_variance8x8_mmx;
if (flags & HAS_SSE2) vp8_variance8x8 = vp8_variance8x8_wmt;
vp8_variance_halfpixvar16x16_h = vp8_variance_halfpixvar16x16_h_c;
if (flags & HAS_MMX) vp8_variance_halfpixvar16x16_h = vp8_variance_halfpixvar16x16_h_mmx;
if (flags & HAS_SSE2) vp8_variance_halfpixvar16x16_h = vp8_variance_halfpixvar16x16_h_wmt;

214
media/libvpx/vp8_rtcd_x86-win32-gcc.h
Просмотреть файл

@ -74,10 +74,10 @@ void vp8_clear_system_state_c();
void vpx_reset_mmx_state();
RTCD_EXTERN void (*vp8_clear_system_state)();
void vp8_copy32xn_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, int n);
void vp8_copy32xn_sse2(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, int n);
void vp8_copy32xn_sse3(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, int n);
RTCD_EXTERN void (*vp8_copy32xn)(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, int n);
void vp8_copy32xn_c(const unsigned char *src_ptr, int source_stride, unsigned char *dst_ptr, int dst_stride, int n);
void vp8_copy32xn_sse2(const unsigned char *src_ptr, int source_stride, unsigned char *dst_ptr, int dst_stride, int n);
void vp8_copy32xn_sse3(const unsigned char *src_ptr, int source_stride, unsigned char *dst_ptr, int dst_stride, int n);
RTCD_EXTERN void (*vp8_copy32xn)(const unsigned char *src_ptr, int source_stride, unsigned char *dst_ptr, int dst_stride, int n);
void vp8_copy_mem16x16_c(unsigned char *src, int src_pitch, unsigned char *dst, int dst_pitch);
void vp8_copy_mem16x16_mmx(unsigned char *src, int src_pitch, unsigned char *dst, int dst_pitch);
@ -147,15 +147,6 @@ int vp8_full_search_sadx3(struct macroblock *x, struct block *b, struct blockd *
int vp8_full_search_sadx8(struct macroblock *x, struct block *b, struct blockd *d, union int_mv *ref_mv, int sad_per_bit, int distance, struct variance_vtable *fn_ptr, int *mvcost[2], union int_mv *center_mv);
RTCD_EXTERN int (*vp8_full_search_sad)(struct macroblock *x, struct block *b, struct blockd *d, union int_mv *ref_mv, int sad_per_bit, int distance, struct variance_vtable *fn_ptr, int *mvcost[2], union int_mv *center_mv);
unsigned int vp8_get4x4sse_cs_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride);
unsigned int vp8_get4x4sse_cs_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride);
RTCD_EXTERN unsigned int (*vp8_get4x4sse_cs)(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride);
unsigned int vp8_get_mb_ss_c(const short *);
unsigned int vp8_get_mb_ss_mmx(const short *);
unsigned int vp8_get_mb_ss_sse2(const short *);
RTCD_EXTERN unsigned int (*vp8_get_mb_ss)(const short *);
void vp8_intra4x4_predict_c(unsigned char *Above, unsigned char *yleft, int left_stride, int b_mode, unsigned char *dst, int dst_stride, unsigned char top_left);
#define vp8_intra4x4_predict vp8_intra4x4_predict_c
@ -218,11 +209,6 @@ int vp8_mbuverror_mmx(struct macroblock *mb);
int vp8_mbuverror_xmm(struct macroblock *mb);
RTCD_EXTERN int (*vp8_mbuverror)(struct macroblock *mb);
unsigned int vp8_mse16x16_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_mse16x16_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_mse16x16_wmt(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
RTCD_EXTERN unsigned int (*vp8_mse16x16)(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
void vp8_plane_add_noise_c(unsigned char *s, char *noise, char blackclamp[16], char whiteclamp[16], char bothclamp[16], unsigned int w, unsigned int h, int pitch);
void vp8_plane_add_noise_mmx(unsigned char *s, char *noise, char blackclamp[16], char whiteclamp[16], char bothclamp[16], unsigned int w, unsigned int h, int pitch);
void vp8_plane_add_noise_wmt(unsigned char *s, char *noise, char blackclamp[16], char whiteclamp[16], char bothclamp[16], unsigned int w, unsigned int h, int pitch);
@ -241,94 +227,6 @@ void vp8_regular_quantize_b_sse2(struct block *, struct blockd *);
void vp8_regular_quantize_b_sse4_1(struct block *, struct blockd *);
RTCD_EXTERN void (*vp8_regular_quantize_b)(struct block *, struct blockd *);
unsigned int vp8_sad16x16_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad16x16_mmx(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad16x16_wmt(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad16x16_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
RTCD_EXTERN unsigned int (*vp8_sad16x16)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
void vp8_sad16x16x3_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad16x16x3_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad16x16x3_ssse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad16x16x3)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad16x16x4d_c(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad16x16x4d_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad16x16x4d)(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad16x16x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
void vp8_sad16x16x8_sse4(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
RTCD_EXTERN void (*vp8_sad16x16x8)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
unsigned int vp8_sad16x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad16x8_mmx(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad16x8_wmt(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
RTCD_EXTERN unsigned int (*vp8_sad16x8)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
void vp8_sad16x8x3_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad16x8x3_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad16x8x3_ssse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad16x8x3)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad16x8x4d_c(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad16x8x4d_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad16x8x4d)(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad16x8x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
void vp8_sad16x8x8_sse4(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
RTCD_EXTERN void (*vp8_sad16x8x8)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
unsigned int vp8_sad4x4_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad4x4_mmx(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad4x4_wmt(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
RTCD_EXTERN unsigned int (*vp8_sad4x4)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
void vp8_sad4x4x3_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad4x4x3_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad4x4x3)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad4x4x4d_c(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad4x4x4d_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad4x4x4d)(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad4x4x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
void vp8_sad4x4x8_sse4(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
RTCD_EXTERN void (*vp8_sad4x4x8)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
unsigned int vp8_sad8x16_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad8x16_mmx(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad8x16_wmt(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
RTCD_EXTERN unsigned int (*vp8_sad8x16)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
void vp8_sad8x16x3_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad8x16x3_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad8x16x3)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad8x16x4d_c(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad8x16x4d_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad8x16x4d)(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad8x16x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
void vp8_sad8x16x8_sse4(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
RTCD_EXTERN void (*vp8_sad8x16x8)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
unsigned int vp8_sad8x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad8x8_mmx(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad8x8_wmt(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
RTCD_EXTERN unsigned int (*vp8_sad8x8)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
void vp8_sad8x8x3_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad8x8x3_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad8x8x3)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad8x8x4d_c(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad8x8x4d_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad8x8x4d)(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad8x8x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
void vp8_sad8x8x8_sse4(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
RTCD_EXTERN void (*vp8_sad8x8x8)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
void vp8_short_fdct4x4_c(short *input, short *output, int pitch);
void vp8_short_fdct4x4_mmx(short *input, short *output, int pitch);
void vp8_short_fdct4x4_sse2(short *input, short *output, int pitch);
@ -378,11 +276,6 @@ void vp8_sixtap_predict8x8_sse2(unsigned char *src, int src_pitch, int xofst, in
void vp8_sixtap_predict8x8_ssse3(unsigned char *src, int src_pitch, int xofst, int yofst, unsigned char *dst, int dst_pitch);
RTCD_EXTERN void (*vp8_sixtap_predict8x8)(unsigned char *src, int src_pitch, int xofst, int yofst, unsigned char *dst, int dst_pitch);
unsigned int vp8_sub_pixel_mse16x16_c(const unsigned char *src_ptr, int source_stride, int xoffset, int yoffset, const unsigned char *ref_ptr, int Refstride, unsigned int *sse);
unsigned int vp8_sub_pixel_mse16x16_mmx(const unsigned char *src_ptr, int source_stride, int xoffset, int yoffset, const unsigned char *ref_ptr, int Refstride, unsigned int *sse);
unsigned int vp8_sub_pixel_mse16x16_wmt(const unsigned char *src_ptr, int source_stride, int xoffset, int yoffset, const unsigned char *ref_ptr, int Refstride, unsigned int *sse);
RTCD_EXTERN unsigned int (*vp8_sub_pixel_mse16x16)(const unsigned char *src_ptr, int source_stride, int xoffset, int yoffset, const unsigned char *ref_ptr, int Refstride, unsigned int *sse);
unsigned int vp8_sub_pixel_variance16x16_c(const unsigned char *src_ptr, int source_stride, int xoffset, int yoffset, const unsigned char *ref_ptr, int Refstride, unsigned int *sse);
unsigned int vp8_sub_pixel_variance16x16_mmx(const unsigned char *src_ptr, int source_stride, int xoffset, int yoffset, const unsigned char *ref_ptr, int Refstride, unsigned int *sse);
unsigned int vp8_sub_pixel_variance16x16_wmt(const unsigned char *src_ptr, int source_stride, int xoffset, int yoffset, const unsigned char *ref_ptr, int Refstride, unsigned int *sse);
@ -429,31 +322,6 @@ void vp8_temporal_filter_apply_c(unsigned char *frame1, unsigned int stride, uns
void vp8_temporal_filter_apply_sse2(unsigned char *frame1, unsigned int stride, unsigned char *frame2, unsigned int block_size, int strength, int filter_weight, unsigned int *accumulator, unsigned short *count);
RTCD_EXTERN void (*vp8_temporal_filter_apply)(unsigned char *frame1, unsigned int stride, unsigned char *frame2, unsigned int block_size, int strength, int filter_weight, unsigned int *accumulator, unsigned short *count);
unsigned int vp8_variance16x16_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance16x16_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance16x16_wmt(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
RTCD_EXTERN unsigned int (*vp8_variance16x16)(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance16x8_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance16x8_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance16x8_wmt(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
RTCD_EXTERN unsigned int (*vp8_variance16x8)(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance4x4_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance4x4_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance4x4_wmt(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
RTCD_EXTERN unsigned int (*vp8_variance4x4)(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance8x16_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance8x16_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance8x16_wmt(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
RTCD_EXTERN unsigned int (*vp8_variance8x16)(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance8x8_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance8x8_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance8x8_wmt(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
RTCD_EXTERN unsigned int (*vp8_variance8x8)(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance_halfpixvar16x16_h_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance_halfpixvar16x16_h_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance_halfpixvar16x16_h_wmt(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
@ -540,11 +408,6 @@ static void setup_rtcd_internal(void)
vp8_full_search_sad = vp8_full_search_sad_c;
if (flags & HAS_SSE3) vp8_full_search_sad = vp8_full_search_sadx3;
if (flags & HAS_SSE4_1) vp8_full_search_sad = vp8_full_search_sadx8;
vp8_get4x4sse_cs = vp8_get4x4sse_cs_c;
if (flags & HAS_MMX) vp8_get4x4sse_cs = vp8_get4x4sse_cs_mmx;
vp8_get_mb_ss = vp8_get_mb_ss_c;
if (flags & HAS_MMX) vp8_get_mb_ss = vp8_get_mb_ss_mmx;
if (flags & HAS_SSE2) vp8_get_mb_ss = vp8_get_mb_ss_sse2;
vp8_loop_filter_bh = vp8_loop_filter_bh_c;
if (flags & HAS_MMX) vp8_loop_filter_bh = vp8_loop_filter_bh_mmx;
if (flags & HAS_SSE2) vp8_loop_filter_bh = vp8_loop_filter_bh_sse2;
@ -580,9 +443,6 @@ static void setup_rtcd_internal(void)
vp8_mbuverror = vp8_mbuverror_c;
if (flags & HAS_MMX) vp8_mbuverror = vp8_mbuverror_mmx;
if (flags & HAS_SSE2) vp8_mbuverror = vp8_mbuverror_xmm;
vp8_mse16x16 = vp8_mse16x16_c;
if (flags & HAS_MMX) vp8_mse16x16 = vp8_mse16x16_mmx;
if (flags & HAS_SSE2) vp8_mse16x16 = vp8_mse16x16_wmt;
vp8_plane_add_noise = vp8_plane_add_noise_c;
if (flags & HAS_MMX) vp8_plane_add_noise = vp8_plane_add_noise_mmx;
if (flags & HAS_SSE2) vp8_plane_add_noise = vp8_plane_add_noise_wmt;
@ -593,54 +453,6 @@ static void setup_rtcd_internal(void)
vp8_regular_quantize_b = vp8_regular_quantize_b_c;
if (flags & HAS_SSE2) vp8_regular_quantize_b = vp8_regular_quantize_b_sse2;
if (flags & HAS_SSE4_1) vp8_regular_quantize_b = vp8_regular_quantize_b_sse4_1;
vp8_sad16x16 = vp8_sad16x16_c;
if (flags & HAS_MMX) vp8_sad16x16 = vp8_sad16x16_mmx;
if (flags & HAS_SSE2) vp8_sad16x16 = vp8_sad16x16_wmt;
if (flags & HAS_SSE3) vp8_sad16x16 = vp8_sad16x16_sse3;
vp8_sad16x16x3 = vp8_sad16x16x3_c;
if (flags & HAS_SSE3) vp8_sad16x16x3 = vp8_sad16x16x3_sse3;
if (flags & HAS_SSSE3) vp8_sad16x16x3 = vp8_sad16x16x3_ssse3;
vp8_sad16x16x4d = vp8_sad16x16x4d_c;
if (flags & HAS_SSE3) vp8_sad16x16x4d = vp8_sad16x16x4d_sse3;
vp8_sad16x16x8 = vp8_sad16x16x8_c;
if (flags & HAS_SSE4_1) vp8_sad16x16x8 = vp8_sad16x16x8_sse4;
vp8_sad16x8 = vp8_sad16x8_c;
if (flags & HAS_MMX) vp8_sad16x8 = vp8_sad16x8_mmx;
if (flags & HAS_SSE2) vp8_sad16x8 = vp8_sad16x8_wmt;
vp8_sad16x8x3 = vp8_sad16x8x3_c;
if (flags & HAS_SSE3) vp8_sad16x8x3 = vp8_sad16x8x3_sse3;
if (flags & HAS_SSSE3) vp8_sad16x8x3 = vp8_sad16x8x3_ssse3;
vp8_sad16x8x4d = vp8_sad16x8x4d_c;
if (flags & HAS_SSE3) vp8_sad16x8x4d = vp8_sad16x8x4d_sse3;
vp8_sad16x8x8 = vp8_sad16x8x8_c;
if (flags & HAS_SSE4_1) vp8_sad16x8x8 = vp8_sad16x8x8_sse4;
vp8_sad4x4 = vp8_sad4x4_c;
if (flags & HAS_MMX) vp8_sad4x4 = vp8_sad4x4_mmx;
if (flags & HAS_SSE2) vp8_sad4x4 = vp8_sad4x4_wmt;
vp8_sad4x4x3 = vp8_sad4x4x3_c;
if (flags & HAS_SSE3) vp8_sad4x4x3 = vp8_sad4x4x3_sse3;
vp8_sad4x4x4d = vp8_sad4x4x4d_c;
if (flags & HAS_SSE3) vp8_sad4x4x4d = vp8_sad4x4x4d_sse3;
vp8_sad4x4x8 = vp8_sad4x4x8_c;
if (flags & HAS_SSE4_1) vp8_sad4x4x8 = vp8_sad4x4x8_sse4;
vp8_sad8x16 = vp8_sad8x16_c;
if (flags & HAS_MMX) vp8_sad8x16 = vp8_sad8x16_mmx;
if (flags & HAS_SSE2) vp8_sad8x16 = vp8_sad8x16_wmt;
vp8_sad8x16x3 = vp8_sad8x16x3_c;
if (flags & HAS_SSE3) vp8_sad8x16x3 = vp8_sad8x16x3_sse3;
vp8_sad8x16x4d = vp8_sad8x16x4d_c;
if (flags & HAS_SSE3) vp8_sad8x16x4d = vp8_sad8x16x4d_sse3;
vp8_sad8x16x8 = vp8_sad8x16x8_c;
if (flags & HAS_SSE4_1) vp8_sad8x16x8 = vp8_sad8x16x8_sse4;
vp8_sad8x8 = vp8_sad8x8_c;
if (flags & HAS_MMX) vp8_sad8x8 = vp8_sad8x8_mmx;
if (flags & HAS_SSE2) vp8_sad8x8 = vp8_sad8x8_wmt;
vp8_sad8x8x3 = vp8_sad8x8x3_c;
if (flags & HAS_SSE3) vp8_sad8x8x3 = vp8_sad8x8x3_sse3;
vp8_sad8x8x4d = vp8_sad8x8x4d_c;
if (flags & HAS_SSE3) vp8_sad8x8x4d = vp8_sad8x8x4d_sse3;
vp8_sad8x8x8 = vp8_sad8x8x8_c;
if (flags & HAS_SSE4_1) vp8_sad8x8x8 = vp8_sad8x8x8_sse4;
vp8_short_fdct4x4 = vp8_short_fdct4x4_c;
if (flags & HAS_MMX) vp8_short_fdct4x4 = vp8_short_fdct4x4_mmx;
if (flags & HAS_SSE2) vp8_short_fdct4x4 = vp8_short_fdct4x4_sse2;
@ -669,9 +481,6 @@ static void setup_rtcd_internal(void)
if (flags & HAS_MMX) vp8_sixtap_predict8x8 = vp8_sixtap_predict8x8_mmx;
if (flags & HAS_SSE2) vp8_sixtap_predict8x8 = vp8_sixtap_predict8x8_sse2;
if (flags & HAS_SSSE3) vp8_sixtap_predict8x8 = vp8_sixtap_predict8x8_ssse3;
vp8_sub_pixel_mse16x16 = vp8_sub_pixel_mse16x16_c;
if (flags & HAS_MMX) vp8_sub_pixel_mse16x16 = vp8_sub_pixel_mse16x16_mmx;
if (flags & HAS_SSE2) vp8_sub_pixel_mse16x16 = vp8_sub_pixel_mse16x16_wmt;
vp8_sub_pixel_variance16x16 = vp8_sub_pixel_variance16x16_c;
if (flags & HAS_MMX) vp8_sub_pixel_variance16x16 = vp8_sub_pixel_variance16x16_mmx;
if (flags & HAS_SSE2) vp8_sub_pixel_variance16x16 = vp8_sub_pixel_variance16x16_wmt;
@ -700,21 +509,6 @@ static void setup_rtcd_internal(void)
if (flags & HAS_SSE2) vp8_subtract_mby = vp8_subtract_mby_sse2;
vp8_temporal_filter_apply = vp8_temporal_filter_apply_c;
if (flags & HAS_SSE2) vp8_temporal_filter_apply = vp8_temporal_filter_apply_sse2;
vp8_variance16x16 = vp8_variance16x16_c;
if (flags & HAS_MMX) vp8_variance16x16 = vp8_variance16x16_mmx;
if (flags & HAS_SSE2) vp8_variance16x16 = vp8_variance16x16_wmt;
vp8_variance16x8 = vp8_variance16x8_c;
if (flags & HAS_MMX) vp8_variance16x8 = vp8_variance16x8_mmx;
if (flags & HAS_SSE2) vp8_variance16x8 = vp8_variance16x8_wmt;
vp8_variance4x4 = vp8_variance4x4_c;
if (flags & HAS_MMX) vp8_variance4x4 = vp8_variance4x4_mmx;
if (flags & HAS_SSE2) vp8_variance4x4 = vp8_variance4x4_wmt;
vp8_variance8x16 = vp8_variance8x16_c;
if (flags & HAS_MMX) vp8_variance8x16 = vp8_variance8x16_mmx;
if (flags & HAS_SSE2) vp8_variance8x16 = vp8_variance8x16_wmt;
vp8_variance8x8 = vp8_variance8x8_c;
if (flags & HAS_MMX) vp8_variance8x8 = vp8_variance8x8_mmx;
if (flags & HAS_SSE2) vp8_variance8x8 = vp8_variance8x8_wmt;
vp8_variance_halfpixvar16x16_h = vp8_variance_halfpixvar16x16_h_c;
if (flags & HAS_MMX) vp8_variance_halfpixvar16x16_h = vp8_variance_halfpixvar16x16_h_mmx;
if (flags & HAS_SSE2) vp8_variance_halfpixvar16x16_h = vp8_variance_halfpixvar16x16_h_wmt;

214
media/libvpx/vp8_rtcd_x86-win32-vs12.h
Просмотреть файл

@ -74,10 +74,10 @@ void vp8_clear_system_state_c();
void vpx_reset_mmx_state();
RTCD_EXTERN void (*vp8_clear_system_state)();
void vp8_copy32xn_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, int n);
void vp8_copy32xn_sse2(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, int n);
void vp8_copy32xn_sse3(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, int n);
RTCD_EXTERN void (*vp8_copy32xn)(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, int n);
void vp8_copy32xn_c(const unsigned char *src_ptr, int source_stride, unsigned char *dst_ptr, int dst_stride, int n);
void vp8_copy32xn_sse2(const unsigned char *src_ptr, int source_stride, unsigned char *dst_ptr, int dst_stride, int n);
void vp8_copy32xn_sse3(const unsigned char *src_ptr, int source_stride, unsigned char *dst_ptr, int dst_stride, int n);
RTCD_EXTERN void (*vp8_copy32xn)(const unsigned char *src_ptr, int source_stride, unsigned char *dst_ptr, int dst_stride, int n);
void vp8_copy_mem16x16_c(unsigned char *src, int src_pitch, unsigned char *dst, int dst_pitch);
void vp8_copy_mem16x16_mmx(unsigned char *src, int src_pitch, unsigned char *dst, int dst_pitch);
@ -147,15 +147,6 @@ int vp8_full_search_sadx3(struct macroblock *x, struct block *b, struct blockd *
int vp8_full_search_sadx8(struct macroblock *x, struct block *b, struct blockd *d, union int_mv *ref_mv, int sad_per_bit, int distance, struct variance_vtable *fn_ptr, int *mvcost[2], union int_mv *center_mv);
RTCD_EXTERN int (*vp8_full_search_sad)(struct macroblock *x, struct block *b, struct blockd *d, union int_mv *ref_mv, int sad_per_bit, int distance, struct variance_vtable *fn_ptr, int *mvcost[2], union int_mv *center_mv);
unsigned int vp8_get4x4sse_cs_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride);
unsigned int vp8_get4x4sse_cs_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride);
RTCD_EXTERN unsigned int (*vp8_get4x4sse_cs)(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride);
unsigned int vp8_get_mb_ss_c(const short *);
unsigned int vp8_get_mb_ss_mmx(const short *);
unsigned int vp8_get_mb_ss_sse2(const short *);
RTCD_EXTERN unsigned int (*vp8_get_mb_ss)(const short *);
void vp8_intra4x4_predict_c(unsigned char *Above, unsigned char *yleft, int left_stride, int b_mode, unsigned char *dst, int dst_stride, unsigned char top_left);
#define vp8_intra4x4_predict vp8_intra4x4_predict_c
@ -218,11 +209,6 @@ int vp8_mbuverror_mmx(struct macroblock *mb);
int vp8_mbuverror_xmm(struct macroblock *mb);
RTCD_EXTERN int (*vp8_mbuverror)(struct macroblock *mb);
unsigned int vp8_mse16x16_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_mse16x16_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_mse16x16_wmt(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
RTCD_EXTERN unsigned int (*vp8_mse16x16)(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
void vp8_plane_add_noise_c(unsigned char *s, char *noise, char blackclamp[16], char whiteclamp[16], char bothclamp[16], unsigned int w, unsigned int h, int pitch);
void vp8_plane_add_noise_mmx(unsigned char *s, char *noise, char blackclamp[16], char whiteclamp[16], char bothclamp[16], unsigned int w, unsigned int h, int pitch);
void vp8_plane_add_noise_wmt(unsigned char *s, char *noise, char blackclamp[16], char whiteclamp[16], char bothclamp[16], unsigned int w, unsigned int h, int pitch);
@ -241,94 +227,6 @@ void vp8_regular_quantize_b_sse2(struct block *, struct blockd *);
void vp8_regular_quantize_b_sse4_1(struct block *, struct blockd *);
RTCD_EXTERN void (*vp8_regular_quantize_b)(struct block *, struct blockd *);
unsigned int vp8_sad16x16_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad16x16_mmx(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad16x16_wmt(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad16x16_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
RTCD_EXTERN unsigned int (*vp8_sad16x16)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
void vp8_sad16x16x3_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad16x16x3_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad16x16x3_ssse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad16x16x3)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad16x16x4d_c(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad16x16x4d_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad16x16x4d)(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad16x16x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
void vp8_sad16x16x8_sse4(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
RTCD_EXTERN void (*vp8_sad16x16x8)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
unsigned int vp8_sad16x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad16x8_mmx(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad16x8_wmt(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
RTCD_EXTERN unsigned int (*vp8_sad16x8)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
void vp8_sad16x8x3_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad16x8x3_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad16x8x3_ssse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad16x8x3)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad16x8x4d_c(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad16x8x4d_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad16x8x4d)(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad16x8x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
void vp8_sad16x8x8_sse4(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
RTCD_EXTERN void (*vp8_sad16x8x8)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
unsigned int vp8_sad4x4_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad4x4_mmx(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad4x4_wmt(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
RTCD_EXTERN unsigned int (*vp8_sad4x4)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
void vp8_sad4x4x3_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad4x4x3_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad4x4x3)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad4x4x4d_c(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad4x4x4d_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad4x4x4d)(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad4x4x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
void vp8_sad4x4x8_sse4(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
RTCD_EXTERN void (*vp8_sad4x4x8)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
unsigned int vp8_sad8x16_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad8x16_mmx(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad8x16_wmt(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
RTCD_EXTERN unsigned int (*vp8_sad8x16)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
void vp8_sad8x16x3_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad8x16x3_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad8x16x3)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad8x16x4d_c(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad8x16x4d_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad8x16x4d)(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad8x16x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
void vp8_sad8x16x8_sse4(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
RTCD_EXTERN void (*vp8_sad8x16x8)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
unsigned int vp8_sad8x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad8x8_mmx(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad8x8_wmt(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
RTCD_EXTERN unsigned int (*vp8_sad8x8)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
void vp8_sad8x8x3_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad8x8x3_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad8x8x3)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad8x8x4d_c(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad8x8x4d_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad8x8x4d)(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad8x8x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
void vp8_sad8x8x8_sse4(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
RTCD_EXTERN void (*vp8_sad8x8x8)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
void vp8_short_fdct4x4_c(short *input, short *output, int pitch);
void vp8_short_fdct4x4_mmx(short *input, short *output, int pitch);
void vp8_short_fdct4x4_sse2(short *input, short *output, int pitch);
@ -378,11 +276,6 @@ void vp8_sixtap_predict8x8_sse2(unsigned char *src, int src_pitch, int xofst, in
void vp8_sixtap_predict8x8_ssse3(unsigned char *src, int src_pitch, int xofst, int yofst, unsigned char *dst, int dst_pitch);
RTCD_EXTERN void (*vp8_sixtap_predict8x8)(unsigned char *src, int src_pitch, int xofst, int yofst, unsigned char *dst, int dst_pitch);
unsigned int vp8_sub_pixel_mse16x16_c(const unsigned char *src_ptr, int source_stride, int xoffset, int yoffset, const unsigned char *ref_ptr, int Refstride, unsigned int *sse);
unsigned int vp8_sub_pixel_mse16x16_mmx(const unsigned char *src_ptr, int source_stride, int xoffset, int yoffset, const unsigned char *ref_ptr, int Refstride, unsigned int *sse);
unsigned int vp8_sub_pixel_mse16x16_wmt(const unsigned char *src_ptr, int source_stride, int xoffset, int yoffset, const unsigned char *ref_ptr, int Refstride, unsigned int *sse);
RTCD_EXTERN unsigned int (*vp8_sub_pixel_mse16x16)(const unsigned char *src_ptr, int source_stride, int xoffset, int yoffset, const unsigned char *ref_ptr, int Refstride, unsigned int *sse);
unsigned int vp8_sub_pixel_variance16x16_c(const unsigned char *src_ptr, int source_stride, int xoffset, int yoffset, const unsigned char *ref_ptr, int Refstride, unsigned int *sse);
unsigned int vp8_sub_pixel_variance16x16_mmx(const unsigned char *src_ptr, int source_stride, int xoffset, int yoffset, const unsigned char *ref_ptr, int Refstride, unsigned int *sse);
unsigned int vp8_sub_pixel_variance16x16_wmt(const unsigned char *src_ptr, int source_stride, int xoffset, int yoffset, const unsigned char *ref_ptr, int Refstride, unsigned int *sse);
@ -429,31 +322,6 @@ void vp8_temporal_filter_apply_c(unsigned char *frame1, unsigned int stride, uns
void vp8_temporal_filter_apply_sse2(unsigned char *frame1, unsigned int stride, unsigned char *frame2, unsigned int block_size, int strength, int filter_weight, unsigned int *accumulator, unsigned short *count);
RTCD_EXTERN void (*vp8_temporal_filter_apply)(unsigned char *frame1, unsigned int stride, unsigned char *frame2, unsigned int block_size, int strength, int filter_weight, unsigned int *accumulator, unsigned short *count);
unsigned int vp8_variance16x16_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance16x16_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance16x16_wmt(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
RTCD_EXTERN unsigned int (*vp8_variance16x16)(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance16x8_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance16x8_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance16x8_wmt(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
RTCD_EXTERN unsigned int (*vp8_variance16x8)(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance4x4_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance4x4_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance4x4_wmt(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
RTCD_EXTERN unsigned int (*vp8_variance4x4)(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance8x16_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance8x16_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance8x16_wmt(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
RTCD_EXTERN unsigned int (*vp8_variance8x16)(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance8x8_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance8x8_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance8x8_wmt(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
RTCD_EXTERN unsigned int (*vp8_variance8x8)(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance_halfpixvar16x16_h_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance_halfpixvar16x16_h_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance_halfpixvar16x16_h_wmt(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
@ -540,11 +408,6 @@ static void setup_rtcd_internal(void)
vp8_full_search_sad = vp8_full_search_sad_c;
if (flags & HAS_SSE3) vp8_full_search_sad = vp8_full_search_sadx3;
if (flags & HAS_SSE4_1) vp8_full_search_sad = vp8_full_search_sadx8;
vp8_get4x4sse_cs = vp8_get4x4sse_cs_c;
if (flags & HAS_MMX) vp8_get4x4sse_cs = vp8_get4x4sse_cs_mmx;
vp8_get_mb_ss = vp8_get_mb_ss_c;
if (flags & HAS_MMX) vp8_get_mb_ss = vp8_get_mb_ss_mmx;
if (flags & HAS_SSE2) vp8_get_mb_ss = vp8_get_mb_ss_sse2;
vp8_loop_filter_bh = vp8_loop_filter_bh_c;
if (flags & HAS_MMX) vp8_loop_filter_bh = vp8_loop_filter_bh_mmx;
if (flags & HAS_SSE2) vp8_loop_filter_bh = vp8_loop_filter_bh_sse2;
@ -580,9 +443,6 @@ static void setup_rtcd_internal(void)
vp8_mbuverror = vp8_mbuverror_c;
if (flags & HAS_MMX) vp8_mbuverror = vp8_mbuverror_mmx;
if (flags & HAS_SSE2) vp8_mbuverror = vp8_mbuverror_xmm;
vp8_mse16x16 = vp8_mse16x16_c;
if (flags & HAS_MMX) vp8_mse16x16 = vp8_mse16x16_mmx;
if (flags & HAS_SSE2) vp8_mse16x16 = vp8_mse16x16_wmt;
vp8_plane_add_noise = vp8_plane_add_noise_c;
if (flags & HAS_MMX) vp8_plane_add_noise = vp8_plane_add_noise_mmx;
if (flags & HAS_SSE2) vp8_plane_add_noise = vp8_plane_add_noise_wmt;
@ -593,54 +453,6 @@ static void setup_rtcd_internal(void)
vp8_regular_quantize_b = vp8_regular_quantize_b_c;
if (flags & HAS_SSE2) vp8_regular_quantize_b = vp8_regular_quantize_b_sse2;
if (flags & HAS_SSE4_1) vp8_regular_quantize_b = vp8_regular_quantize_b_sse4_1;
vp8_sad16x16 = vp8_sad16x16_c;
if (flags & HAS_MMX) vp8_sad16x16 = vp8_sad16x16_mmx;
if (flags & HAS_SSE2) vp8_sad16x16 = vp8_sad16x16_wmt;
if (flags & HAS_SSE3) vp8_sad16x16 = vp8_sad16x16_sse3;
vp8_sad16x16x3 = vp8_sad16x16x3_c;
if (flags & HAS_SSE3) vp8_sad16x16x3 = vp8_sad16x16x3_sse3;
if (flags & HAS_SSSE3) vp8_sad16x16x3 = vp8_sad16x16x3_ssse3;
vp8_sad16x16x4d = vp8_sad16x16x4d_c;
if (flags & HAS_SSE3) vp8_sad16x16x4d = vp8_sad16x16x4d_sse3;
vp8_sad16x16x8 = vp8_sad16x16x8_c;
if (flags & HAS_SSE4_1) vp8_sad16x16x8 = vp8_sad16x16x8_sse4;
vp8_sad16x8 = vp8_sad16x8_c;
if (flags & HAS_MMX) vp8_sad16x8 = vp8_sad16x8_mmx;
if (flags & HAS_SSE2) vp8_sad16x8 = vp8_sad16x8_wmt;
vp8_sad16x8x3 = vp8_sad16x8x3_c;
if (flags & HAS_SSE3) vp8_sad16x8x3 = vp8_sad16x8x3_sse3;
if (flags & HAS_SSSE3) vp8_sad16x8x3 = vp8_sad16x8x3_ssse3;
vp8_sad16x8x4d = vp8_sad16x8x4d_c;
if (flags & HAS_SSE3) vp8_sad16x8x4d = vp8_sad16x8x4d_sse3;
vp8_sad16x8x8 = vp8_sad16x8x8_c;
if (flags & HAS_SSE4_1) vp8_sad16x8x8 = vp8_sad16x8x8_sse4;
vp8_sad4x4 = vp8_sad4x4_c;
if (flags & HAS_MMX) vp8_sad4x4 = vp8_sad4x4_mmx;
if (flags & HAS_SSE2) vp8_sad4x4 = vp8_sad4x4_wmt;
vp8_sad4x4x3 = vp8_sad4x4x3_c;
if (flags & HAS_SSE3) vp8_sad4x4x3 = vp8_sad4x4x3_sse3;
vp8_sad4x4x4d = vp8_sad4x4x4d_c;
if (flags & HAS_SSE3) vp8_sad4x4x4d = vp8_sad4x4x4d_sse3;
vp8_sad4x4x8 = vp8_sad4x4x8_c;
if (flags & HAS_SSE4_1) vp8_sad4x4x8 = vp8_sad4x4x8_sse4;
vp8_sad8x16 = vp8_sad8x16_c;
if (flags & HAS_MMX) vp8_sad8x16 = vp8_sad8x16_mmx;
if (flags & HAS_SSE2) vp8_sad8x16 = vp8_sad8x16_wmt;
vp8_sad8x16x3 = vp8_sad8x16x3_c;
if (flags & HAS_SSE3) vp8_sad8x16x3 = vp8_sad8x16x3_sse3;
vp8_sad8x16x4d = vp8_sad8x16x4d_c;
if (flags & HAS_SSE3) vp8_sad8x16x4d = vp8_sad8x16x4d_sse3;
vp8_sad8x16x8 = vp8_sad8x16x8_c;
if (flags & HAS_SSE4_1) vp8_sad8x16x8 = vp8_sad8x16x8_sse4;
vp8_sad8x8 = vp8_sad8x8_c;
if (flags & HAS_MMX) vp8_sad8x8 = vp8_sad8x8_mmx;
if (flags & HAS_SSE2) vp8_sad8x8 = vp8_sad8x8_wmt;
vp8_sad8x8x3 = vp8_sad8x8x3_c;
if (flags & HAS_SSE3) vp8_sad8x8x3 = vp8_sad8x8x3_sse3;
vp8_sad8x8x4d = vp8_sad8x8x4d_c;
if (flags & HAS_SSE3) vp8_sad8x8x4d = vp8_sad8x8x4d_sse3;
vp8_sad8x8x8 = vp8_sad8x8x8_c;
if (flags & HAS_SSE4_1) vp8_sad8x8x8 = vp8_sad8x8x8_sse4;
vp8_short_fdct4x4 = vp8_short_fdct4x4_c;
if (flags & HAS_MMX) vp8_short_fdct4x4 = vp8_short_fdct4x4_mmx;
if (flags & HAS_SSE2) vp8_short_fdct4x4 = vp8_short_fdct4x4_sse2;
@ -669,9 +481,6 @@ static void setup_rtcd_internal(void)
if (flags & HAS_MMX) vp8_sixtap_predict8x8 = vp8_sixtap_predict8x8_mmx;
if (flags & HAS_SSE2) vp8_sixtap_predict8x8 = vp8_sixtap_predict8x8_sse2;
if (flags & HAS_SSSE3) vp8_sixtap_predict8x8 = vp8_sixtap_predict8x8_ssse3;
vp8_sub_pixel_mse16x16 = vp8_sub_pixel_mse16x16_c;
if (flags & HAS_MMX) vp8_sub_pixel_mse16x16 = vp8_sub_pixel_mse16x16_mmx;
if (flags & HAS_SSE2) vp8_sub_pixel_mse16x16 = vp8_sub_pixel_mse16x16_wmt;
vp8_sub_pixel_variance16x16 = vp8_sub_pixel_variance16x16_c;
if (flags & HAS_MMX) vp8_sub_pixel_variance16x16 = vp8_sub_pixel_variance16x16_mmx;
if (flags & HAS_SSE2) vp8_sub_pixel_variance16x16 = vp8_sub_pixel_variance16x16_wmt;
@ -700,21 +509,6 @@ static void setup_rtcd_internal(void)
if (flags & HAS_SSE2) vp8_subtract_mby = vp8_subtract_mby_sse2;
vp8_temporal_filter_apply = vp8_temporal_filter_apply_c;
if (flags & HAS_SSE2) vp8_temporal_filter_apply = vp8_temporal_filter_apply_sse2;
vp8_variance16x16 = vp8_variance16x16_c;
if (flags & HAS_MMX) vp8_variance16x16 = vp8_variance16x16_mmx;
if (flags & HAS_SSE2) vp8_variance16x16 = vp8_variance16x16_wmt;
vp8_variance16x8 = vp8_variance16x8_c;
if (flags & HAS_MMX) vp8_variance16x8 = vp8_variance16x8_mmx;
if (flags & HAS_SSE2) vp8_variance16x8 = vp8_variance16x8_wmt;
vp8_variance4x4 = vp8_variance4x4_c;
if (flags & HAS_MMX) vp8_variance4x4 = vp8_variance4x4_mmx;
if (flags & HAS_SSE2) vp8_variance4x4 = vp8_variance4x4_wmt;
vp8_variance8x16 = vp8_variance8x16_c;
if (flags & HAS_MMX) vp8_variance8x16 = vp8_variance8x16_mmx;
if (flags & HAS_SSE2) vp8_variance8x16 = vp8_variance8x16_wmt;
vp8_variance8x8 = vp8_variance8x8_c;
if (flags & HAS_MMX) vp8_variance8x8 = vp8_variance8x8_mmx;
if (flags & HAS_SSE2) vp8_variance8x8 = vp8_variance8x8_wmt;
vp8_variance_halfpixvar16x16_h = vp8_variance_halfpixvar16x16_h_c;
if (flags & HAS_MMX) vp8_variance_halfpixvar16x16_h = vp8_variance_halfpixvar16x16_h_mmx;
if (flags & HAS_SSE2) vp8_variance_halfpixvar16x16_h = vp8_variance_halfpixvar16x16_h_wmt;

174
media/libvpx/vp8_rtcd_x86_64-darwin9-gcc.h
Просмотреть файл

@ -74,10 +74,10 @@ void vp8_clear_system_state_c();
void vpx_reset_mmx_state();
#define vp8_clear_system_state vpx_reset_mmx_state
void vp8_copy32xn_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, int n);
void vp8_copy32xn_sse2(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, int n);
void vp8_copy32xn_sse3(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, int n);
RTCD_EXTERN void (*vp8_copy32xn)(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, int n);
void vp8_copy32xn_c(const unsigned char *src_ptr, int source_stride, unsigned char *dst_ptr, int dst_stride, int n);
void vp8_copy32xn_sse2(const unsigned char *src_ptr, int source_stride, unsigned char *dst_ptr, int dst_stride, int n);
void vp8_copy32xn_sse3(const unsigned char *src_ptr, int source_stride, unsigned char *dst_ptr, int dst_stride, int n);
RTCD_EXTERN void (*vp8_copy32xn)(const unsigned char *src_ptr, int source_stride, unsigned char *dst_ptr, int dst_stride, int n);
void vp8_copy_mem16x16_c(unsigned char *src, int src_pitch, unsigned char *dst, int dst_pitch);
void vp8_copy_mem16x16_mmx(unsigned char *src, int src_pitch, unsigned char *dst, int dst_pitch);
@ -147,15 +147,6 @@ int vp8_full_search_sadx3(struct macroblock *x, struct block *b, struct blockd *
int vp8_full_search_sadx8(struct macroblock *x, struct block *b, struct blockd *d, union int_mv *ref_mv, int sad_per_bit, int distance, struct variance_vtable *fn_ptr, int *mvcost[2], union int_mv *center_mv);
RTCD_EXTERN int (*vp8_full_search_sad)(struct macroblock *x, struct block *b, struct blockd *d, union int_mv *ref_mv, int sad_per_bit, int distance, struct variance_vtable *fn_ptr, int *mvcost[2], union int_mv *center_mv);
unsigned int vp8_get4x4sse_cs_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride);
unsigned int vp8_get4x4sse_cs_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride);
#define vp8_get4x4sse_cs vp8_get4x4sse_cs_mmx
unsigned int vp8_get_mb_ss_c(const short *);
unsigned int vp8_get_mb_ss_mmx(const short *);
unsigned int vp8_get_mb_ss_sse2(const short *);
#define vp8_get_mb_ss vp8_get_mb_ss_sse2
void vp8_intra4x4_predict_c(unsigned char *Above, unsigned char *yleft, int left_stride, int b_mode, unsigned char *dst, int dst_stride, unsigned char top_left);
#define vp8_intra4x4_predict vp8_intra4x4_predict_c
@ -218,11 +209,6 @@ int vp8_mbuverror_mmx(struct macroblock *mb);
int vp8_mbuverror_xmm(struct macroblock *mb);
#define vp8_mbuverror vp8_mbuverror_xmm
unsigned int vp8_mse16x16_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_mse16x16_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_mse16x16_wmt(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
#define vp8_mse16x16 vp8_mse16x16_wmt
void vp8_plane_add_noise_c(unsigned char *s, char *noise, char blackclamp[16], char whiteclamp[16], char bothclamp[16], unsigned int w, unsigned int h, int pitch);
void vp8_plane_add_noise_mmx(unsigned char *s, char *noise, char blackclamp[16], char whiteclamp[16], char bothclamp[16], unsigned int w, unsigned int h, int pitch);
void vp8_plane_add_noise_wmt(unsigned char *s, char *noise, char blackclamp[16], char whiteclamp[16], char bothclamp[16], unsigned int w, unsigned int h, int pitch);
@ -241,94 +227,6 @@ void vp8_regular_quantize_b_sse2(struct block *, struct blockd *);
void vp8_regular_quantize_b_sse4_1(struct block *, struct blockd *);
RTCD_EXTERN void (*vp8_regular_quantize_b)(struct block *, struct blockd *);
unsigned int vp8_sad16x16_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad16x16_mmx(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad16x16_wmt(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad16x16_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
RTCD_EXTERN unsigned int (*vp8_sad16x16)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
void vp8_sad16x16x3_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad16x16x3_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad16x16x3_ssse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad16x16x3)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad16x16x4d_c(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad16x16x4d_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad16x16x4d)(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad16x16x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
void vp8_sad16x16x8_sse4(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
RTCD_EXTERN void (*vp8_sad16x16x8)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
unsigned int vp8_sad16x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad16x8_mmx(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad16x8_wmt(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
#define vp8_sad16x8 vp8_sad16x8_wmt
void vp8_sad16x8x3_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad16x8x3_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad16x8x3_ssse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad16x8x3)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad16x8x4d_c(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad16x8x4d_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad16x8x4d)(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad16x8x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
void vp8_sad16x8x8_sse4(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
RTCD_EXTERN void (*vp8_sad16x8x8)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
unsigned int vp8_sad4x4_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad4x4_mmx(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad4x4_wmt(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
#define vp8_sad4x4 vp8_sad4x4_wmt
void vp8_sad4x4x3_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad4x4x3_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad4x4x3)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad4x4x4d_c(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad4x4x4d_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad4x4x4d)(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad4x4x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
void vp8_sad4x4x8_sse4(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
RTCD_EXTERN void (*vp8_sad4x4x8)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
unsigned int vp8_sad8x16_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad8x16_mmx(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad8x16_wmt(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
#define vp8_sad8x16 vp8_sad8x16_wmt
void vp8_sad8x16x3_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad8x16x3_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad8x16x3)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad8x16x4d_c(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad8x16x4d_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad8x16x4d)(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad8x16x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
void vp8_sad8x16x8_sse4(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
RTCD_EXTERN void (*vp8_sad8x16x8)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
unsigned int vp8_sad8x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad8x8_mmx(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad8x8_wmt(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
#define vp8_sad8x8 vp8_sad8x8_wmt
void vp8_sad8x8x3_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad8x8x3_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad8x8x3)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad8x8x4d_c(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad8x8x4d_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad8x8x4d)(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad8x8x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
void vp8_sad8x8x8_sse4(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
RTCD_EXTERN void (*vp8_sad8x8x8)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
void vp8_short_fdct4x4_c(short *input, short *output, int pitch);
void vp8_short_fdct4x4_mmx(short *input, short *output, int pitch);
void vp8_short_fdct4x4_sse2(short *input, short *output, int pitch);
@ -378,11 +276,6 @@ void vp8_sixtap_predict8x8_sse2(unsigned char *src, int src_pitch, int xofst, in
void vp8_sixtap_predict8x8_ssse3(unsigned char *src, int src_pitch, int xofst, int yofst, unsigned char *dst, int dst_pitch);
RTCD_EXTERN void (*vp8_sixtap_predict8x8)(unsigned char *src, int src_pitch, int xofst, int yofst, unsigned char *dst, int dst_pitch);
unsigned int vp8_sub_pixel_mse16x16_c(const unsigned char *src_ptr, int source_stride, int xoffset, int yoffset, const unsigned char *ref_ptr, int Refstride, unsigned int *sse);
unsigned int vp8_sub_pixel_mse16x16_mmx(const unsigned char *src_ptr, int source_stride, int xoffset, int yoffset, const unsigned char *ref_ptr, int Refstride, unsigned int *sse);
unsigned int vp8_sub_pixel_mse16x16_wmt(const unsigned char *src_ptr, int source_stride, int xoffset, int yoffset, const unsigned char *ref_ptr, int Refstride, unsigned int *sse);
#define vp8_sub_pixel_mse16x16 vp8_sub_pixel_mse16x16_wmt
unsigned int vp8_sub_pixel_variance16x16_c(const unsigned char *src_ptr, int source_stride, int xoffset, int yoffset, const unsigned char *ref_ptr, int Refstride, unsigned int *sse);
unsigned int vp8_sub_pixel_variance16x16_mmx(const unsigned char *src_ptr, int source_stride, int xoffset, int yoffset, const unsigned char *ref_ptr, int Refstride, unsigned int *sse);
unsigned int vp8_sub_pixel_variance16x16_wmt(const unsigned char *src_ptr, int source_stride, int xoffset, int yoffset, const unsigned char *ref_ptr, int Refstride, unsigned int *sse);
@ -429,31 +322,6 @@ void vp8_temporal_filter_apply_c(unsigned char *frame1, unsigned int stride, uns
void vp8_temporal_filter_apply_sse2(unsigned char *frame1, unsigned int stride, unsigned char *frame2, unsigned int block_size, int strength, int filter_weight, unsigned int *accumulator, unsigned short *count);
#define vp8_temporal_filter_apply vp8_temporal_filter_apply_sse2
unsigned int vp8_variance16x16_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance16x16_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance16x16_wmt(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
#define vp8_variance16x16 vp8_variance16x16_wmt
unsigned int vp8_variance16x8_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance16x8_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance16x8_wmt(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
#define vp8_variance16x8 vp8_variance16x8_wmt
unsigned int vp8_variance4x4_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance4x4_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance4x4_wmt(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
#define vp8_variance4x4 vp8_variance4x4_wmt
unsigned int vp8_variance8x16_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance8x16_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance8x16_wmt(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
#define vp8_variance8x16 vp8_variance8x16_wmt
unsigned int vp8_variance8x8_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance8x8_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance8x8_wmt(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
#define vp8_variance8x8 vp8_variance8x8_wmt
unsigned int vp8_variance_halfpixvar16x16_h_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance_halfpixvar16x16_h_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance_halfpixvar16x16_h_wmt(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
@ -500,40 +368,6 @@ static void setup_rtcd_internal(void)
if (flags & HAS_SSE3) vp8_refining_search_sad = vp8_refining_search_sadx4;
vp8_regular_quantize_b = vp8_regular_quantize_b_sse2;
if (flags & HAS_SSE4_1) vp8_regular_quantize_b = vp8_regular_quantize_b_sse4_1;
vp8_sad16x16 = vp8_sad16x16_wmt;
if (flags & HAS_SSE3) vp8_sad16x16 = vp8_sad16x16_sse3;
vp8_sad16x16x3 = vp8_sad16x16x3_c;
if (flags & HAS_SSE3) vp8_sad16x16x3 = vp8_sad16x16x3_sse3;
if (flags & HAS_SSSE3) vp8_sad16x16x3 = vp8_sad16x16x3_ssse3;
vp8_sad16x16x4d = vp8_sad16x16x4d_c;
if (flags & HAS_SSE3) vp8_sad16x16x4d = vp8_sad16x16x4d_sse3;
vp8_sad16x16x8 = vp8_sad16x16x8_c;
if (flags & HAS_SSE4_1) vp8_sad16x16x8 = vp8_sad16x16x8_sse4;
vp8_sad16x8x3 = vp8_sad16x8x3_c;
if (flags & HAS_SSE3) vp8_sad16x8x3 = vp8_sad16x8x3_sse3;
if (flags & HAS_SSSE3) vp8_sad16x8x3 = vp8_sad16x8x3_ssse3;
vp8_sad16x8x4d = vp8_sad16x8x4d_c;
if (flags & HAS_SSE3) vp8_sad16x8x4d = vp8_sad16x8x4d_sse3;
vp8_sad16x8x8 = vp8_sad16x8x8_c;
if (flags & HAS_SSE4_1) vp8_sad16x8x8 = vp8_sad16x8x8_sse4;
vp8_sad4x4x3 = vp8_sad4x4x3_c;
if (flags & HAS_SSE3) vp8_sad4x4x3 = vp8_sad4x4x3_sse3;
vp8_sad4x4x4d = vp8_sad4x4x4d_c;
if (flags & HAS_SSE3) vp8_sad4x4x4d = vp8_sad4x4x4d_sse3;
vp8_sad4x4x8 = vp8_sad4x4x8_c;
if (flags & HAS_SSE4_1) vp8_sad4x4x8 = vp8_sad4x4x8_sse4;
vp8_sad8x16x3 = vp8_sad8x16x3_c;
if (flags & HAS_SSE3) vp8_sad8x16x3 = vp8_sad8x16x3_sse3;
vp8_sad8x16x4d = vp8_sad8x16x4d_c;
if (flags & HAS_SSE3) vp8_sad8x16x4d = vp8_sad8x16x4d_sse3;
vp8_sad8x16x8 = vp8_sad8x16x8_c;
if (flags & HAS_SSE4_1) vp8_sad8x16x8 = vp8_sad8x16x8_sse4;
vp8_sad8x8x3 = vp8_sad8x8x3_c;
if (flags & HAS_SSE3) vp8_sad8x8x3 = vp8_sad8x8x3_sse3;
vp8_sad8x8x4d = vp8_sad8x8x4d_c;
if (flags & HAS_SSE3) vp8_sad8x8x4d = vp8_sad8x8x4d_sse3;
vp8_sad8x8x8 = vp8_sad8x8x8_c;
if (flags & HAS_SSE4_1) vp8_sad8x8x8 = vp8_sad8x8x8_sse4;
vp8_sixtap_predict16x16 = vp8_sixtap_predict16x16_sse2;
if (flags & HAS_SSSE3) vp8_sixtap_predict16x16 = vp8_sixtap_predict16x16_ssse3;
vp8_sixtap_predict4x4 = vp8_sixtap_predict4x4_mmx;

174
media/libvpx/vp8_rtcd_x86_64-linux-gcc.h
Просмотреть файл

@ -74,10 +74,10 @@ void vp8_clear_system_state_c();
void vpx_reset_mmx_state();
#define vp8_clear_system_state vpx_reset_mmx_state
void vp8_copy32xn_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, int n);
void vp8_copy32xn_sse2(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, int n);
void vp8_copy32xn_sse3(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, int n);
RTCD_EXTERN void (*vp8_copy32xn)(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, int n);
void vp8_copy32xn_c(const unsigned char *src_ptr, int source_stride, unsigned char *dst_ptr, int dst_stride, int n);
void vp8_copy32xn_sse2(const unsigned char *src_ptr, int source_stride, unsigned char *dst_ptr, int dst_stride, int n);
void vp8_copy32xn_sse3(const unsigned char *src_ptr, int source_stride, unsigned char *dst_ptr, int dst_stride, int n);
RTCD_EXTERN void (*vp8_copy32xn)(const unsigned char *src_ptr, int source_stride, unsigned char *dst_ptr, int dst_stride, int n);
void vp8_copy_mem16x16_c(unsigned char *src, int src_pitch, unsigned char *dst, int dst_pitch);
void vp8_copy_mem16x16_mmx(unsigned char *src, int src_pitch, unsigned char *dst, int dst_pitch);
@ -147,15 +147,6 @@ int vp8_full_search_sadx3(struct macroblock *x, struct block *b, struct blockd *
int vp8_full_search_sadx8(struct macroblock *x, struct block *b, struct blockd *d, union int_mv *ref_mv, int sad_per_bit, int distance, struct variance_vtable *fn_ptr, int *mvcost[2], union int_mv *center_mv);
RTCD_EXTERN int (*vp8_full_search_sad)(struct macroblock *x, struct block *b, struct blockd *d, union int_mv *ref_mv, int sad_per_bit, int distance, struct variance_vtable *fn_ptr, int *mvcost[2], union int_mv *center_mv);
unsigned int vp8_get4x4sse_cs_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride);
unsigned int vp8_get4x4sse_cs_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride);
#define vp8_get4x4sse_cs vp8_get4x4sse_cs_mmx
unsigned int vp8_get_mb_ss_c(const short *);
unsigned int vp8_get_mb_ss_mmx(const short *);
unsigned int vp8_get_mb_ss_sse2(const short *);
#define vp8_get_mb_ss vp8_get_mb_ss_sse2
void vp8_intra4x4_predict_c(unsigned char *Above, unsigned char *yleft, int left_stride, int b_mode, unsigned char *dst, int dst_stride, unsigned char top_left);
#define vp8_intra4x4_predict vp8_intra4x4_predict_c
@ -218,11 +209,6 @@ int vp8_mbuverror_mmx(struct macroblock *mb);
int vp8_mbuverror_xmm(struct macroblock *mb);
#define vp8_mbuverror vp8_mbuverror_xmm
unsigned int vp8_mse16x16_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_mse16x16_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_mse16x16_wmt(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
#define vp8_mse16x16 vp8_mse16x16_wmt
void vp8_plane_add_noise_c(unsigned char *s, char *noise, char blackclamp[16], char whiteclamp[16], char bothclamp[16], unsigned int w, unsigned int h, int pitch);
void vp8_plane_add_noise_mmx(unsigned char *s, char *noise, char blackclamp[16], char whiteclamp[16], char bothclamp[16], unsigned int w, unsigned int h, int pitch);
void vp8_plane_add_noise_wmt(unsigned char *s, char *noise, char blackclamp[16], char whiteclamp[16], char bothclamp[16], unsigned int w, unsigned int h, int pitch);
@ -241,94 +227,6 @@ void vp8_regular_quantize_b_sse2(struct block *, struct blockd *);
void vp8_regular_quantize_b_sse4_1(struct block *, struct blockd *);
RTCD_EXTERN void (*vp8_regular_quantize_b)(struct block *, struct blockd *);
unsigned int vp8_sad16x16_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad16x16_mmx(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad16x16_wmt(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad16x16_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
RTCD_EXTERN unsigned int (*vp8_sad16x16)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
void vp8_sad16x16x3_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad16x16x3_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad16x16x3_ssse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad16x16x3)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad16x16x4d_c(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad16x16x4d_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad16x16x4d)(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad16x16x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
void vp8_sad16x16x8_sse4(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
RTCD_EXTERN void (*vp8_sad16x16x8)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
unsigned int vp8_sad16x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad16x8_mmx(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad16x8_wmt(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
#define vp8_sad16x8 vp8_sad16x8_wmt
void vp8_sad16x8x3_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad16x8x3_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad16x8x3_ssse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad16x8x3)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad16x8x4d_c(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad16x8x4d_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad16x8x4d)(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad16x8x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
void vp8_sad16x8x8_sse4(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
RTCD_EXTERN void (*vp8_sad16x8x8)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
unsigned int vp8_sad4x4_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad4x4_mmx(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad4x4_wmt(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
#define vp8_sad4x4 vp8_sad4x4_wmt
void vp8_sad4x4x3_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad4x4x3_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad4x4x3)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad4x4x4d_c(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad4x4x4d_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad4x4x4d)(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad4x4x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
void vp8_sad4x4x8_sse4(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
RTCD_EXTERN void (*vp8_sad4x4x8)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
unsigned int vp8_sad8x16_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad8x16_mmx(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad8x16_wmt(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
#define vp8_sad8x16 vp8_sad8x16_wmt
void vp8_sad8x16x3_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad8x16x3_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad8x16x3)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad8x16x4d_c(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad8x16x4d_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad8x16x4d)(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad8x16x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
void vp8_sad8x16x8_sse4(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
RTCD_EXTERN void (*vp8_sad8x16x8)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
unsigned int vp8_sad8x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad8x8_mmx(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad8x8_wmt(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
#define vp8_sad8x8 vp8_sad8x8_wmt
void vp8_sad8x8x3_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad8x8x3_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad8x8x3)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad8x8x4d_c(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad8x8x4d_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad8x8x4d)(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad8x8x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
void vp8_sad8x8x8_sse4(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
RTCD_EXTERN void (*vp8_sad8x8x8)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
void vp8_short_fdct4x4_c(short *input, short *output, int pitch);
void vp8_short_fdct4x4_mmx(short *input, short *output, int pitch);
void vp8_short_fdct4x4_sse2(short *input, short *output, int pitch);
@ -378,11 +276,6 @@ void vp8_sixtap_predict8x8_sse2(unsigned char *src, int src_pitch, int xofst, in
void vp8_sixtap_predict8x8_ssse3(unsigned char *src, int src_pitch, int xofst, int yofst, unsigned char *dst, int dst_pitch);
RTCD_EXTERN void (*vp8_sixtap_predict8x8)(unsigned char *src, int src_pitch, int xofst, int yofst, unsigned char *dst, int dst_pitch);
unsigned int vp8_sub_pixel_mse16x16_c(const unsigned char *src_ptr, int source_stride, int xoffset, int yoffset, const unsigned char *ref_ptr, int Refstride, unsigned int *sse);
unsigned int vp8_sub_pixel_mse16x16_mmx(const unsigned char *src_ptr, int source_stride, int xoffset, int yoffset, const unsigned char *ref_ptr, int Refstride, unsigned int *sse);
unsigned int vp8_sub_pixel_mse16x16_wmt(const unsigned char *src_ptr, int source_stride, int xoffset, int yoffset, const unsigned char *ref_ptr, int Refstride, unsigned int *sse);
#define vp8_sub_pixel_mse16x16 vp8_sub_pixel_mse16x16_wmt
unsigned int vp8_sub_pixel_variance16x16_c(const unsigned char *src_ptr, int source_stride, int xoffset, int yoffset, const unsigned char *ref_ptr, int Refstride, unsigned int *sse);
unsigned int vp8_sub_pixel_variance16x16_mmx(const unsigned char *src_ptr, int source_stride, int xoffset, int yoffset, const unsigned char *ref_ptr, int Refstride, unsigned int *sse);
unsigned int vp8_sub_pixel_variance16x16_wmt(const unsigned char *src_ptr, int source_stride, int xoffset, int yoffset, const unsigned char *ref_ptr, int Refstride, unsigned int *sse);
@ -429,31 +322,6 @@ void vp8_temporal_filter_apply_c(unsigned char *frame1, unsigned int stride, uns
void vp8_temporal_filter_apply_sse2(unsigned char *frame1, unsigned int stride, unsigned char *frame2, unsigned int block_size, int strength, int filter_weight, unsigned int *accumulator, unsigned short *count);
#define vp8_temporal_filter_apply vp8_temporal_filter_apply_sse2
unsigned int vp8_variance16x16_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance16x16_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance16x16_wmt(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
#define vp8_variance16x16 vp8_variance16x16_wmt
unsigned int vp8_variance16x8_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance16x8_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance16x8_wmt(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
#define vp8_variance16x8 vp8_variance16x8_wmt
unsigned int vp8_variance4x4_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance4x4_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance4x4_wmt(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
#define vp8_variance4x4 vp8_variance4x4_wmt
unsigned int vp8_variance8x16_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance8x16_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance8x16_wmt(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
#define vp8_variance8x16 vp8_variance8x16_wmt
unsigned int vp8_variance8x8_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance8x8_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance8x8_wmt(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
#define vp8_variance8x8 vp8_variance8x8_wmt
unsigned int vp8_variance_halfpixvar16x16_h_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance_halfpixvar16x16_h_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance_halfpixvar16x16_h_wmt(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
@ -500,40 +368,6 @@ static void setup_rtcd_internal(void)
if (flags & HAS_SSE3) vp8_refining_search_sad = vp8_refining_search_sadx4;
vp8_regular_quantize_b = vp8_regular_quantize_b_sse2;
if (flags & HAS_SSE4_1) vp8_regular_quantize_b = vp8_regular_quantize_b_sse4_1;
vp8_sad16x16 = vp8_sad16x16_wmt;
if (flags & HAS_SSE3) vp8_sad16x16 = vp8_sad16x16_sse3;
vp8_sad16x16x3 = vp8_sad16x16x3_c;
if (flags & HAS_SSE3) vp8_sad16x16x3 = vp8_sad16x16x3_sse3;
if (flags & HAS_SSSE3) vp8_sad16x16x3 = vp8_sad16x16x3_ssse3;
vp8_sad16x16x4d = vp8_sad16x16x4d_c;
if (flags & HAS_SSE3) vp8_sad16x16x4d = vp8_sad16x16x4d_sse3;
vp8_sad16x16x8 = vp8_sad16x16x8_c;
if (flags & HAS_SSE4_1) vp8_sad16x16x8 = vp8_sad16x16x8_sse4;
vp8_sad16x8x3 = vp8_sad16x8x3_c;
if (flags & HAS_SSE3) vp8_sad16x8x3 = vp8_sad16x8x3_sse3;
if (flags & HAS_SSSE3) vp8_sad16x8x3 = vp8_sad16x8x3_ssse3;
vp8_sad16x8x4d = vp8_sad16x8x4d_c;
if (flags & HAS_SSE3) vp8_sad16x8x4d = vp8_sad16x8x4d_sse3;
vp8_sad16x8x8 = vp8_sad16x8x8_c;
if (flags & HAS_SSE4_1) vp8_sad16x8x8 = vp8_sad16x8x8_sse4;
vp8_sad4x4x3 = vp8_sad4x4x3_c;
if (flags & HAS_SSE3) vp8_sad4x4x3 = vp8_sad4x4x3_sse3;
vp8_sad4x4x4d = vp8_sad4x4x4d_c;
if (flags & HAS_SSE3) vp8_sad4x4x4d = vp8_sad4x4x4d_sse3;
vp8_sad4x4x8 = vp8_sad4x4x8_c;
if (flags & HAS_SSE4_1) vp8_sad4x4x8 = vp8_sad4x4x8_sse4;
vp8_sad8x16x3 = vp8_sad8x16x3_c;
if (flags & HAS_SSE3) vp8_sad8x16x3 = vp8_sad8x16x3_sse3;
vp8_sad8x16x4d = vp8_sad8x16x4d_c;
if (flags & HAS_SSE3) vp8_sad8x16x4d = vp8_sad8x16x4d_sse3;
vp8_sad8x16x8 = vp8_sad8x16x8_c;
if (flags & HAS_SSE4_1) vp8_sad8x16x8 = vp8_sad8x16x8_sse4;
vp8_sad8x8x3 = vp8_sad8x8x3_c;
if (flags & HAS_SSE3) vp8_sad8x8x3 = vp8_sad8x8x3_sse3;
vp8_sad8x8x4d = vp8_sad8x8x4d_c;
if (flags & HAS_SSE3) vp8_sad8x8x4d = vp8_sad8x8x4d_sse3;
vp8_sad8x8x8 = vp8_sad8x8x8_c;
if (flags & HAS_SSE4_1) vp8_sad8x8x8 = vp8_sad8x8x8_sse4;
vp8_sixtap_predict16x16 = vp8_sixtap_predict16x16_sse2;
if (flags & HAS_SSSE3) vp8_sixtap_predict16x16 = vp8_sixtap_predict16x16_ssse3;
vp8_sixtap_predict4x4 = vp8_sixtap_predict4x4_mmx;

174
media/libvpx/vp8_rtcd_x86_64-win64-gcc.h
Просмотреть файл

@ -74,10 +74,10 @@ void vp8_clear_system_state_c();
void vpx_reset_mmx_state();
#define vp8_clear_system_state vpx_reset_mmx_state
void vp8_copy32xn_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, int n);
void vp8_copy32xn_sse2(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, int n);
void vp8_copy32xn_sse3(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, int n);
RTCD_EXTERN void (*vp8_copy32xn)(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, int n);
void vp8_copy32xn_c(const unsigned char *src_ptr, int source_stride, unsigned char *dst_ptr, int dst_stride, int n);
void vp8_copy32xn_sse2(const unsigned char *src_ptr, int source_stride, unsigned char *dst_ptr, int dst_stride, int n);
void vp8_copy32xn_sse3(const unsigned char *src_ptr, int source_stride, unsigned char *dst_ptr, int dst_stride, int n);
RTCD_EXTERN void (*vp8_copy32xn)(const unsigned char *src_ptr, int source_stride, unsigned char *dst_ptr, int dst_stride, int n);
void vp8_copy_mem16x16_c(unsigned char *src, int src_pitch, unsigned char *dst, int dst_pitch);
void vp8_copy_mem16x16_mmx(unsigned char *src, int src_pitch, unsigned char *dst, int dst_pitch);
@ -147,15 +147,6 @@ int vp8_full_search_sadx3(struct macroblock *x, struct block *b, struct blockd *
int vp8_full_search_sadx8(struct macroblock *x, struct block *b, struct blockd *d, union int_mv *ref_mv, int sad_per_bit, int distance, struct variance_vtable *fn_ptr, int *mvcost[2], union int_mv *center_mv);
RTCD_EXTERN int (*vp8_full_search_sad)(struct macroblock *x, struct block *b, struct blockd *d, union int_mv *ref_mv, int sad_per_bit, int distance, struct variance_vtable *fn_ptr, int *mvcost[2], union int_mv *center_mv);
unsigned int vp8_get4x4sse_cs_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride);
unsigned int vp8_get4x4sse_cs_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride);
#define vp8_get4x4sse_cs vp8_get4x4sse_cs_mmx
unsigned int vp8_get_mb_ss_c(const short *);
unsigned int vp8_get_mb_ss_mmx(const short *);
unsigned int vp8_get_mb_ss_sse2(const short *);
#define vp8_get_mb_ss vp8_get_mb_ss_sse2
void vp8_intra4x4_predict_c(unsigned char *Above, unsigned char *yleft, int left_stride, int b_mode, unsigned char *dst, int dst_stride, unsigned char top_left);
#define vp8_intra4x4_predict vp8_intra4x4_predict_c
@ -218,11 +209,6 @@ int vp8_mbuverror_mmx(struct macroblock *mb);
int vp8_mbuverror_xmm(struct macroblock *mb);
#define vp8_mbuverror vp8_mbuverror_xmm
unsigned int vp8_mse16x16_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_mse16x16_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_mse16x16_wmt(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
#define vp8_mse16x16 vp8_mse16x16_wmt
void vp8_plane_add_noise_c(unsigned char *s, char *noise, char blackclamp[16], char whiteclamp[16], char bothclamp[16], unsigned int w, unsigned int h, int pitch);
void vp8_plane_add_noise_mmx(unsigned char *s, char *noise, char blackclamp[16], char whiteclamp[16], char bothclamp[16], unsigned int w, unsigned int h, int pitch);
void vp8_plane_add_noise_wmt(unsigned char *s, char *noise, char blackclamp[16], char whiteclamp[16], char bothclamp[16], unsigned int w, unsigned int h, int pitch);
@ -241,94 +227,6 @@ void vp8_regular_quantize_b_sse2(struct block *, struct blockd *);
void vp8_regular_quantize_b_sse4_1(struct block *, struct blockd *);
RTCD_EXTERN void (*vp8_regular_quantize_b)(struct block *, struct blockd *);
unsigned int vp8_sad16x16_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad16x16_mmx(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad16x16_wmt(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad16x16_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
RTCD_EXTERN unsigned int (*vp8_sad16x16)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
void vp8_sad16x16x3_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad16x16x3_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad16x16x3_ssse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad16x16x3)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad16x16x4d_c(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad16x16x4d_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad16x16x4d)(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad16x16x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
void vp8_sad16x16x8_sse4(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
RTCD_EXTERN void (*vp8_sad16x16x8)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
unsigned int vp8_sad16x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad16x8_mmx(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad16x8_wmt(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
#define vp8_sad16x8 vp8_sad16x8_wmt
void vp8_sad16x8x3_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad16x8x3_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad16x8x3_ssse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad16x8x3)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad16x8x4d_c(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad16x8x4d_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad16x8x4d)(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad16x8x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
void vp8_sad16x8x8_sse4(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
RTCD_EXTERN void (*vp8_sad16x8x8)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
unsigned int vp8_sad4x4_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad4x4_mmx(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad4x4_wmt(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
#define vp8_sad4x4 vp8_sad4x4_wmt
void vp8_sad4x4x3_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad4x4x3_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad4x4x3)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad4x4x4d_c(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad4x4x4d_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad4x4x4d)(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad4x4x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
void vp8_sad4x4x8_sse4(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
RTCD_EXTERN void (*vp8_sad4x4x8)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
unsigned int vp8_sad8x16_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad8x16_mmx(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad8x16_wmt(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
#define vp8_sad8x16 vp8_sad8x16_wmt
void vp8_sad8x16x3_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad8x16x3_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad8x16x3)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad8x16x4d_c(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad8x16x4d_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad8x16x4d)(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad8x16x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
void vp8_sad8x16x8_sse4(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
RTCD_EXTERN void (*vp8_sad8x16x8)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
unsigned int vp8_sad8x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad8x8_mmx(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad8x8_wmt(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
#define vp8_sad8x8 vp8_sad8x8_wmt
void vp8_sad8x8x3_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad8x8x3_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad8x8x3)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad8x8x4d_c(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad8x8x4d_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad8x8x4d)(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad8x8x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
void vp8_sad8x8x8_sse4(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
RTCD_EXTERN void (*vp8_sad8x8x8)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
void vp8_short_fdct4x4_c(short *input, short *output, int pitch);
void vp8_short_fdct4x4_mmx(short *input, short *output, int pitch);
void vp8_short_fdct4x4_sse2(short *input, short *output, int pitch);
@ -378,11 +276,6 @@ void vp8_sixtap_predict8x8_sse2(unsigned char *src, int src_pitch, int xofst, in
void vp8_sixtap_predict8x8_ssse3(unsigned char *src, int src_pitch, int xofst, int yofst, unsigned char *dst, int dst_pitch);
RTCD_EXTERN void (*vp8_sixtap_predict8x8)(unsigned char *src, int src_pitch, int xofst, int yofst, unsigned char *dst, int dst_pitch);
unsigned int vp8_sub_pixel_mse16x16_c(const unsigned char *src_ptr, int source_stride, int xoffset, int yoffset, const unsigned char *ref_ptr, int Refstride, unsigned int *sse);
unsigned int vp8_sub_pixel_mse16x16_mmx(const unsigned char *src_ptr, int source_stride, int xoffset, int yoffset, const unsigned char *ref_ptr, int Refstride, unsigned int *sse);
unsigned int vp8_sub_pixel_mse16x16_wmt(const unsigned char *src_ptr, int source_stride, int xoffset, int yoffset, const unsigned char *ref_ptr, int Refstride, unsigned int *sse);
#define vp8_sub_pixel_mse16x16 vp8_sub_pixel_mse16x16_wmt
unsigned int vp8_sub_pixel_variance16x16_c(const unsigned char *src_ptr, int source_stride, int xoffset, int yoffset, const unsigned char *ref_ptr, int Refstride, unsigned int *sse);
unsigned int vp8_sub_pixel_variance16x16_mmx(const unsigned char *src_ptr, int source_stride, int xoffset, int yoffset, const unsigned char *ref_ptr, int Refstride, unsigned int *sse);
unsigned int vp8_sub_pixel_variance16x16_wmt(const unsigned char *src_ptr, int source_stride, int xoffset, int yoffset, const unsigned char *ref_ptr, int Refstride, unsigned int *sse);
@ -429,31 +322,6 @@ void vp8_temporal_filter_apply_c(unsigned char *frame1, unsigned int stride, uns
void vp8_temporal_filter_apply_sse2(unsigned char *frame1, unsigned int stride, unsigned char *frame2, unsigned int block_size, int strength, int filter_weight, unsigned int *accumulator, unsigned short *count);
#define vp8_temporal_filter_apply vp8_temporal_filter_apply_sse2
unsigned int vp8_variance16x16_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance16x16_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance16x16_wmt(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
#define vp8_variance16x16 vp8_variance16x16_wmt
unsigned int vp8_variance16x8_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance16x8_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance16x8_wmt(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
#define vp8_variance16x8 vp8_variance16x8_wmt
unsigned int vp8_variance4x4_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance4x4_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance4x4_wmt(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
#define vp8_variance4x4 vp8_variance4x4_wmt
unsigned int vp8_variance8x16_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance8x16_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance8x16_wmt(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
#define vp8_variance8x16 vp8_variance8x16_wmt
unsigned int vp8_variance8x8_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance8x8_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance8x8_wmt(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
#define vp8_variance8x8 vp8_variance8x8_wmt
unsigned int vp8_variance_halfpixvar16x16_h_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance_halfpixvar16x16_h_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance_halfpixvar16x16_h_wmt(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
@ -500,40 +368,6 @@ static void setup_rtcd_internal(void)
if (flags & HAS_SSE3) vp8_refining_search_sad = vp8_refining_search_sadx4;
vp8_regular_quantize_b = vp8_regular_quantize_b_sse2;
if (flags & HAS_SSE4_1) vp8_regular_quantize_b = vp8_regular_quantize_b_sse4_1;
vp8_sad16x16 = vp8_sad16x16_wmt;
if (flags & HAS_SSE3) vp8_sad16x16 = vp8_sad16x16_sse3;
vp8_sad16x16x3 = vp8_sad16x16x3_c;
if (flags & HAS_SSE3) vp8_sad16x16x3 = vp8_sad16x16x3_sse3;
if (flags & HAS_SSSE3) vp8_sad16x16x3 = vp8_sad16x16x3_ssse3;
vp8_sad16x16x4d = vp8_sad16x16x4d_c;
if (flags & HAS_SSE3) vp8_sad16x16x4d = vp8_sad16x16x4d_sse3;
vp8_sad16x16x8 = vp8_sad16x16x8_c;
if (flags & HAS_SSE4_1) vp8_sad16x16x8 = vp8_sad16x16x8_sse4;
vp8_sad16x8x3 = vp8_sad16x8x3_c;
if (flags & HAS_SSE3) vp8_sad16x8x3 = vp8_sad16x8x3_sse3;
if (flags & HAS_SSSE3) vp8_sad16x8x3 = vp8_sad16x8x3_ssse3;
vp8_sad16x8x4d = vp8_sad16x8x4d_c;
if (flags & HAS_SSE3) vp8_sad16x8x4d = vp8_sad16x8x4d_sse3;
vp8_sad16x8x8 = vp8_sad16x8x8_c;
if (flags & HAS_SSE4_1) vp8_sad16x8x8 = vp8_sad16x8x8_sse4;
vp8_sad4x4x3 = vp8_sad4x4x3_c;
if (flags & HAS_SSE3) vp8_sad4x4x3 = vp8_sad4x4x3_sse3;
vp8_sad4x4x4d = vp8_sad4x4x4d_c;
if (flags & HAS_SSE3) vp8_sad4x4x4d = vp8_sad4x4x4d_sse3;
vp8_sad4x4x8 = vp8_sad4x4x8_c;
if (flags & HAS_SSE4_1) vp8_sad4x4x8 = vp8_sad4x4x8_sse4;
vp8_sad8x16x3 = vp8_sad8x16x3_c;
if (flags & HAS_SSE3) vp8_sad8x16x3 = vp8_sad8x16x3_sse3;
vp8_sad8x16x4d = vp8_sad8x16x4d_c;
if (flags & HAS_SSE3) vp8_sad8x16x4d = vp8_sad8x16x4d_sse3;
vp8_sad8x16x8 = vp8_sad8x16x8_c;
if (flags & HAS_SSE4_1) vp8_sad8x16x8 = vp8_sad8x16x8_sse4;
vp8_sad8x8x3 = vp8_sad8x8x3_c;
if (flags & HAS_SSE3) vp8_sad8x8x3 = vp8_sad8x8x3_sse3;
vp8_sad8x8x4d = vp8_sad8x8x4d_c;
if (flags & HAS_SSE3) vp8_sad8x8x4d = vp8_sad8x8x4d_sse3;
vp8_sad8x8x8 = vp8_sad8x8x8_c;
if (flags & HAS_SSE4_1) vp8_sad8x8x8 = vp8_sad8x8x8_sse4;
vp8_sixtap_predict16x16 = vp8_sixtap_predict16x16_sse2;
if (flags & HAS_SSSE3) vp8_sixtap_predict16x16 = vp8_sixtap_predict16x16_ssse3;
vp8_sixtap_predict4x4 = vp8_sixtap_predict4x4_mmx;

174
media/libvpx/vp8_rtcd_x86_64-win64-vs12.h
Просмотреть файл

@ -74,10 +74,10 @@ void vp8_clear_system_state_c();
void vpx_reset_mmx_state();
#define vp8_clear_system_state vpx_reset_mmx_state
void vp8_copy32xn_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, int n);
void vp8_copy32xn_sse2(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, int n);
void vp8_copy32xn_sse3(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, int n);
RTCD_EXTERN void (*vp8_copy32xn)(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, int n);
void vp8_copy32xn_c(const unsigned char *src_ptr, int source_stride, unsigned char *dst_ptr, int dst_stride, int n);
void vp8_copy32xn_sse2(const unsigned char *src_ptr, int source_stride, unsigned char *dst_ptr, int dst_stride, int n);
void vp8_copy32xn_sse3(const unsigned char *src_ptr, int source_stride, unsigned char *dst_ptr, int dst_stride, int n);
RTCD_EXTERN void (*vp8_copy32xn)(const unsigned char *src_ptr, int source_stride, unsigned char *dst_ptr, int dst_stride, int n);
void vp8_copy_mem16x16_c(unsigned char *src, int src_pitch, unsigned char *dst, int dst_pitch);
void vp8_copy_mem16x16_mmx(unsigned char *src, int src_pitch, unsigned char *dst, int dst_pitch);
@ -147,15 +147,6 @@ int vp8_full_search_sadx3(struct macroblock *x, struct block *b, struct blockd *
int vp8_full_search_sadx8(struct macroblock *x, struct block *b, struct blockd *d, union int_mv *ref_mv, int sad_per_bit, int distance, struct variance_vtable *fn_ptr, int *mvcost[2], union int_mv *center_mv);
RTCD_EXTERN int (*vp8_full_search_sad)(struct macroblock *x, struct block *b, struct blockd *d, union int_mv *ref_mv, int sad_per_bit, int distance, struct variance_vtable *fn_ptr, int *mvcost[2], union int_mv *center_mv);
unsigned int vp8_get4x4sse_cs_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride);
unsigned int vp8_get4x4sse_cs_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride);
#define vp8_get4x4sse_cs vp8_get4x4sse_cs_mmx
unsigned int vp8_get_mb_ss_c(const short *);
unsigned int vp8_get_mb_ss_mmx(const short *);
unsigned int vp8_get_mb_ss_sse2(const short *);
#define vp8_get_mb_ss vp8_get_mb_ss_sse2
void vp8_intra4x4_predict_c(unsigned char *Above, unsigned char *yleft, int left_stride, int b_mode, unsigned char *dst, int dst_stride, unsigned char top_left);
#define vp8_intra4x4_predict vp8_intra4x4_predict_c
@ -218,11 +209,6 @@ int vp8_mbuverror_mmx(struct macroblock *mb);
int vp8_mbuverror_xmm(struct macroblock *mb);
#define vp8_mbuverror vp8_mbuverror_xmm
unsigned int vp8_mse16x16_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_mse16x16_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_mse16x16_wmt(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
#define vp8_mse16x16 vp8_mse16x16_wmt
void vp8_plane_add_noise_c(unsigned char *s, char *noise, char blackclamp[16], char whiteclamp[16], char bothclamp[16], unsigned int w, unsigned int h, int pitch);
void vp8_plane_add_noise_mmx(unsigned char *s, char *noise, char blackclamp[16], char whiteclamp[16], char bothclamp[16], unsigned int w, unsigned int h, int pitch);
void vp8_plane_add_noise_wmt(unsigned char *s, char *noise, char blackclamp[16], char whiteclamp[16], char bothclamp[16], unsigned int w, unsigned int h, int pitch);
@ -241,94 +227,6 @@ void vp8_regular_quantize_b_sse2(struct block *, struct blockd *);
void vp8_regular_quantize_b_sse4_1(struct block *, struct blockd *);
RTCD_EXTERN void (*vp8_regular_quantize_b)(struct block *, struct blockd *);
unsigned int vp8_sad16x16_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad16x16_mmx(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad16x16_wmt(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad16x16_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
RTCD_EXTERN unsigned int (*vp8_sad16x16)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
void vp8_sad16x16x3_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad16x16x3_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad16x16x3_ssse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad16x16x3)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad16x16x4d_c(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad16x16x4d_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad16x16x4d)(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad16x16x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
void vp8_sad16x16x8_sse4(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
RTCD_EXTERN void (*vp8_sad16x16x8)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
unsigned int vp8_sad16x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad16x8_mmx(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad16x8_wmt(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
#define vp8_sad16x8 vp8_sad16x8_wmt
void vp8_sad16x8x3_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad16x8x3_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad16x8x3_ssse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad16x8x3)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad16x8x4d_c(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad16x8x4d_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad16x8x4d)(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad16x8x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
void vp8_sad16x8x8_sse4(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
RTCD_EXTERN void (*vp8_sad16x8x8)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
unsigned int vp8_sad4x4_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad4x4_mmx(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad4x4_wmt(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
#define vp8_sad4x4 vp8_sad4x4_wmt
void vp8_sad4x4x3_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad4x4x3_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad4x4x3)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad4x4x4d_c(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad4x4x4d_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad4x4x4d)(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad4x4x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
void vp8_sad4x4x8_sse4(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
RTCD_EXTERN void (*vp8_sad4x4x8)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
unsigned int vp8_sad8x16_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad8x16_mmx(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad8x16_wmt(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
#define vp8_sad8x16 vp8_sad8x16_wmt
void vp8_sad8x16x3_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad8x16x3_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad8x16x3)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad8x16x4d_c(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad8x16x4d_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad8x16x4d)(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad8x16x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
void vp8_sad8x16x8_sse4(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
RTCD_EXTERN void (*vp8_sad8x16x8)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
unsigned int vp8_sad8x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad8x8_mmx(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
unsigned int vp8_sad8x8_wmt(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad);
#define vp8_sad8x8 vp8_sad8x8_wmt
void vp8_sad8x8x3_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad8x8x3_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad8x8x3)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sad_array);
void vp8_sad8x8x4d_c(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad8x8x4d_sse3(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
RTCD_EXTERN void (*vp8_sad8x8x4d)(const unsigned char *src_ptr, int src_stride, const unsigned char * const ref_ptr[], int ref_stride, unsigned int *sad_array);
void vp8_sad8x8x8_c(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
void vp8_sad8x8x8_sse4(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
RTCD_EXTERN void (*vp8_sad8x8x8)(const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned short *sad_array);
void vp8_short_fdct4x4_c(short *input, short *output, int pitch);
void vp8_short_fdct4x4_mmx(short *input, short *output, int pitch);
void vp8_short_fdct4x4_sse2(short *input, short *output, int pitch);
@ -378,11 +276,6 @@ void vp8_sixtap_predict8x8_sse2(unsigned char *src, int src_pitch, int xofst, in
void vp8_sixtap_predict8x8_ssse3(unsigned char *src, int src_pitch, int xofst, int yofst, unsigned char *dst, int dst_pitch);
RTCD_EXTERN void (*vp8_sixtap_predict8x8)(unsigned char *src, int src_pitch, int xofst, int yofst, unsigned char *dst, int dst_pitch);
unsigned int vp8_sub_pixel_mse16x16_c(const unsigned char *src_ptr, int source_stride, int xoffset, int yoffset, const unsigned char *ref_ptr, int Refstride, unsigned int *sse);
unsigned int vp8_sub_pixel_mse16x16_mmx(const unsigned char *src_ptr, int source_stride, int xoffset, int yoffset, const unsigned char *ref_ptr, int Refstride, unsigned int *sse);
unsigned int vp8_sub_pixel_mse16x16_wmt(const unsigned char *src_ptr, int source_stride, int xoffset, int yoffset, const unsigned char *ref_ptr, int Refstride, unsigned int *sse);
#define vp8_sub_pixel_mse16x16 vp8_sub_pixel_mse16x16_wmt
unsigned int vp8_sub_pixel_variance16x16_c(const unsigned char *src_ptr, int source_stride, int xoffset, int yoffset, const unsigned char *ref_ptr, int Refstride, unsigned int *sse);
unsigned int vp8_sub_pixel_variance16x16_mmx(const unsigned char *src_ptr, int source_stride, int xoffset, int yoffset, const unsigned char *ref_ptr, int Refstride, unsigned int *sse);
unsigned int vp8_sub_pixel_variance16x16_wmt(const unsigned char *src_ptr, int source_stride, int xoffset, int yoffset, const unsigned char *ref_ptr, int Refstride, unsigned int *sse);
@ -429,31 +322,6 @@ void vp8_temporal_filter_apply_c(unsigned char *frame1, unsigned int stride, uns
void vp8_temporal_filter_apply_sse2(unsigned char *frame1, unsigned int stride, unsigned char *frame2, unsigned int block_size, int strength, int filter_weight, unsigned int *accumulator, unsigned short *count);
#define vp8_temporal_filter_apply vp8_temporal_filter_apply_sse2
unsigned int vp8_variance16x16_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance16x16_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance16x16_wmt(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
#define vp8_variance16x16 vp8_variance16x16_wmt
unsigned int vp8_variance16x8_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance16x8_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance16x8_wmt(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
#define vp8_variance16x8 vp8_variance16x8_wmt
unsigned int vp8_variance4x4_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance4x4_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance4x4_wmt(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
#define vp8_variance4x4 vp8_variance4x4_wmt
unsigned int vp8_variance8x16_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance8x16_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance8x16_wmt(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
#define vp8_variance8x16 vp8_variance8x16_wmt
unsigned int vp8_variance8x8_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance8x8_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance8x8_wmt(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
#define vp8_variance8x8 vp8_variance8x8_wmt
unsigned int vp8_variance_halfpixvar16x16_h_c(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance_halfpixvar16x16_h_mmx(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
unsigned int vp8_variance_halfpixvar16x16_h_wmt(const unsigned char *src_ptr, int source_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int *sse);
@ -500,40 +368,6 @@ static void setup_rtcd_internal(void)
if (flags & HAS_SSE3) vp8_refining_search_sad = vp8_refining_search_sadx4;
vp8_regular_quantize_b = vp8_regular_quantize_b_sse2;
if (flags & HAS_SSE4_1) vp8_regular_quantize_b = vp8_regular_quantize_b_sse4_1;
vp8_sad16x16 = vp8_sad16x16_wmt;
if (flags & HAS_SSE3) vp8_sad16x16 = vp8_sad16x16_sse3;
vp8_sad16x16x3 = vp8_sad16x16x3_c;
if (flags & HAS_SSE3) vp8_sad16x16x3 = vp8_sad16x16x3_sse3;
if (flags & HAS_SSSE3) vp8_sad16x16x3 = vp8_sad16x16x3_ssse3;
vp8_sad16x16x4d = vp8_sad16x16x4d_c;
if (flags & HAS_SSE3) vp8_sad16x16x4d = vp8_sad16x16x4d_sse3;
vp8_sad16x16x8 = vp8_sad16x16x8_c;
if (flags & HAS_SSE4_1) vp8_sad16x16x8 = vp8_sad16x16x8_sse4;
vp8_sad16x8x3 = vp8_sad16x8x3_c;
if (flags & HAS_SSE3) vp8_sad16x8x3 = vp8_sad16x8x3_sse3;
if (flags & HAS_SSSE3) vp8_sad16x8x3 = vp8_sad16x8x3_ssse3;
vp8_sad16x8x4d = vp8_sad16x8x4d_c;
if (flags & HAS_SSE3) vp8_sad16x8x4d = vp8_sad16x8x4d_sse3;
vp8_sad16x8x8 = vp8_sad16x8x8_c;
if (flags & HAS_SSE4_1) vp8_sad16x8x8 = vp8_sad16x8x8_sse4;
vp8_sad4x4x3 = vp8_sad4x4x3_c;
if (flags & HAS_SSE3) vp8_sad4x4x3 = vp8_sad4x4x3_sse3;
vp8_sad4x4x4d = vp8_sad4x4x4d_c;
if (flags & HAS_SSE3) vp8_sad4x4x4d = vp8_sad4x4x4d_sse3;
vp8_sad4x4x8 = vp8_sad4x4x8_c;
if (flags & HAS_SSE4_1) vp8_sad4x4x8 = vp8_sad4x4x8_sse4;
vp8_sad8x16x3 = vp8_sad8x16x3_c;
if (flags & HAS_SSE3) vp8_sad8x16x3 = vp8_sad8x16x3_sse3;
vp8_sad8x16x4d = vp8_sad8x16x4d_c;
if (flags & HAS_SSE3) vp8_sad8x16x4d = vp8_sad8x16x4d_sse3;
vp8_sad8x16x8 = vp8_sad8x16x8_c;
if (flags & HAS_SSE4_1) vp8_sad8x16x8 = vp8_sad8x16x8_sse4;
vp8_sad8x8x3 = vp8_sad8x8x3_c;
if (flags & HAS_SSE3) vp8_sad8x8x3 = vp8_sad8x8x3_sse3;
vp8_sad8x8x4d = vp8_sad8x8x4d_c;
if (flags & HAS_SSE3) vp8_sad8x8x4d = vp8_sad8x8x4d_sse3;
vp8_sad8x8x8 = vp8_sad8x8x8_c;
if (flags & HAS_SSE4_1) vp8_sad8x8x8 = vp8_sad8x8x8_sse4;
vp8_sixtap_predict16x16 = vp8_sixtap_predict16x16_sse2;
if (flags & HAS_SSSE3) vp8_sixtap_predict16x16 = vp8_sixtap_predict16x16_ssse3;
vp8_sixtap_predict4x4 = vp8_sixtap_predict4x4_mmx;

0
media/libvpx/vp9/common/arm/neon/vp9_avg_neon.c → media/libvpx/vp9/common/arm/neon/vp9_convolve_avg_neon.c
Просмотреть файл

0
media/libvpx/vp9/common/arm/neon/vp9_avg_neon_asm.asm → media/libvpx/vp9/common/arm/neon/vp9_convolve_avg_neon_asm.asm
Просмотреть файл

4
media/libvpx/vp9/common/arm/neon/vp9_convolve_neon.c
Просмотреть файл

@ -20,7 +20,7 @@ void vp9_convolve8_neon(const uint8_t *src, ptrdiff_t src_stride,
/* Given our constraints: w <= 64, h <= 64, taps == 8 we can reduce the
* maximum buffer size to 64 * 64 + 7 (+ 1 to make it divisible by 4).
*/
DECLARE_ALIGNED_ARRAY(8, uint8_t, temp, 64 * 72);
DECLARE_ALIGNED(8, uint8_t, temp[64 * 72]);
// Account for the vertical phase needing 3 lines prior and 4 lines post
int intermediate_height = h + 7;
@ -56,7 +56,7 @@ void vp9_convolve8_avg_neon(const uint8_t *src, ptrdiff_t src_stride,
const int16_t *filter_x, int x_step_q4,
const int16_t *filter_y, int y_step_q4,
int w, int h) {
DECLARE_ALIGNED_ARRAY(8, uint8_t, temp, 64 * 72);
DECLARE_ALIGNED(8, uint8_t, temp[64 * 72]);
int intermediate_height = h + 7;
if (x_step_q4 != 16 || y_step_q4 != 16) {

2
media/libvpx/vp9/common/arm/neon/vp9_idct16x16_1_add_neon.c
Просмотреть файл

@ -9,6 +9,8 @@
*/
#include <arm_neon.h>
#include "vpx_ports/mem.h"
#include "vp9/common/vp9_idct.h"
void vp9_idct16x16_1_add_neon(

4
media/libvpx/vp9/common/arm/neon/vp9_idct32x32_1_add_neon.c
Просмотреть файл

@ -9,10 +9,12 @@
*/
#include <arm_neon.h>
#include "vp9/common/vp9_idct.h"
#include "./vpx_config.h"
#include "vpx_ports/mem.h"
#include "vp9/common/vp9_idct.h"
static INLINE void LD_16x8(
uint8_t *d,
int d_stride,

2
media/libvpx/vp9/common/arm/neon/vp9_idct4x4_1_add_neon.c
Просмотреть файл

@ -9,6 +9,8 @@
*/
#include <arm_neon.h>
#include "vpx_ports/mem.h"
#include "vp9/common/vp9_idct.h"
void vp9_idct4x4_1_add_neon(

2
media/libvpx/vp9/common/arm/neon/vp9_idct8x8_1_add_neon.c
Просмотреть файл

@ -9,6 +9,8 @@
*/
#include <arm_neon.h>
#include "vpx_ports/mem.h"
#include "vp9/common/vp9_idct.h"
void vp9_idct8x8_1_add_neon(

965
media/libvpx/vp9/common/arm/neon/vp9_reconintra_neon.c
Просмотреть файл

Разница между файлами не показана из-за своего большого размера Загрузить разницу

34
media/libvpx/vp9/common/arm/neon/vp9_reconintra_neon_asm.asm
Просмотреть файл

@ -298,8 +298,7 @@ loop_h
|vp9_tm_predictor_4x4_neon| PROC
; Load ytop_left = above[-1];
sub r12, r2, #1
ldrb r12, [r12]
vdup.u8 d0, r12
vld1.u8 {d0[]}, [r12]
; Load above 4 pixels
vld1.32 {d2[0]}, [r2]
@ -309,10 +308,10 @@ loop_h
; Load left row by row and compute left + (above - ytop_left)
; 1st row and 2nd row
ldrb r12, [r3], #1
ldrb r2, [r3], #1
vdup.u16 q1, r12
vdup.u16 q2, r2
vld1.u8 {d2[]}, [r3]!
vld1.u8 {d4[]}, [r3]!
vmovl.u8 q1, d2
vmovl.u8 q2, d4
vadd.s16 q1, q1, q3
vadd.s16 q2, q2, q3
vqmovun.s16 d0, q1
@ -321,10 +320,10 @@ loop_h
vst1.32 {d1[0]}, [r0], r1
; 3rd row and 4th row
ldrb r12, [r3], #1
ldrb r2, [r3], #1
vdup.u16 q1, r12
vdup.u16 q2, r2
vld1.u8 {d2[]}, [r3]!
vld1.u8 {d4[]}, [r3]
vmovl.u8 q1, d2
vmovl.u8 q2, d4
vadd.s16 q1, q1, q3
vadd.s16 q2, q2, q3
vqmovun.s16 d0, q1
@ -345,8 +344,7 @@ loop_h
|vp9_tm_predictor_8x8_neon| PROC
; Load ytop_left = above[-1];
sub r12, r2, #1
ldrb r12, [r12]
vdup.u8 d0, r12
vld1.8 {d0[]}, [r12]
; preload 8 left
vld1.8 {d30}, [r3]
@ -418,8 +416,7 @@ loop_h
|vp9_tm_predictor_16x16_neon| PROC
; Load ytop_left = above[-1];
sub r12, r2, #1
ldrb r12, [r12]
vdup.u8 q0, r12
vld1.8 {d0[]}, [r12]
; Load above 8 pixels
vld1.8 {q1}, [r2]
@ -429,7 +426,7 @@ loop_h
; Compute above - ytop_left
vsubl.u8 q2, d2, d0
vsubl.u8 q3, d3, d1
vsubl.u8 q3, d3, d0
vmovl.u8 q10, d18
@ -512,8 +509,7 @@ loop_16x16_neon
|vp9_tm_predictor_32x32_neon| PROC
; Load ytop_left = above[-1];
sub r12, r2, #1
ldrb r12, [r12]
vdup.u8 q0, r12
vld1.8 {d0[]}, [r12]
; Load above 32 pixels
vld1.8 {q1}, [r2]!
@ -524,9 +520,9 @@ loop_16x16_neon
; Compute above - ytop_left
vsubl.u8 q8, d2, d0
vsubl.u8 q9, d3, d1
vsubl.u8 q9, d3, d0
vsubl.u8 q10, d4, d0
vsubl.u8 q11, d5, d1
vsubl.u8 q11, d5, d0
vmovl.u8 q3, d26

51
media/libvpx/vp9/common/vp9_alloccommon.c
Просмотреть файл

@ -11,6 +11,7 @@
#include "./vpx_config.h"
#include "vpx_mem/vpx_mem.h"
#include "vp9/common/vp9_alloccommon.h"
#include "vp9/common/vp9_blockd.h"
#include "vp9/common/vp9_entropymode.h"
#include "vp9/common/vp9_entropymv.h"
@ -56,6 +57,7 @@ static int alloc_seg_map(VP9_COMMON *cm, int seg_map_size) {
if (cm->seg_map_array[i] == NULL)
return 1;
}
cm->seg_map_alloc_size = seg_map_size;
// Init the index.
cm->seg_map_idx = 0;
@ -83,8 +85,7 @@ static void free_seg_map(VP9_COMMON *cm) {
}
}
void vp9_free_ref_frame_buffers(VP9_COMMON *cm) {
BufferPool *const pool = cm->buffer_pool;
void vp9_free_ref_frame_buffers(BufferPool *pool) {
int i;
for (i = 0; i < FRAME_BUFFERS; ++i) {
@ -97,10 +98,14 @@ void vp9_free_ref_frame_buffers(VP9_COMMON *cm) {
pool->frame_bufs[i].mvs = NULL;
vp9_free_frame_buffer(&pool->frame_bufs[i].buf);
}
}
void vp9_free_postproc_buffers(VP9_COMMON *cm) {
#if CONFIG_VP9_POSTPROC
vp9_free_frame_buffer(&cm->post_proc_buffer);
vp9_free_frame_buffer(&cm->post_proc_buffer_int);
#else
(void)cm;
#endif
}
@ -114,25 +119,36 @@ void vp9_free_context_buffers(VP9_COMMON *cm) {
}
int vp9_alloc_context_buffers(VP9_COMMON *cm, int width, int height) {
vp9_free_context_buffers(cm);
int new_mi_size;
vp9_set_mb_mi(cm, width, height);
if (cm->alloc_mi(cm, cm->mi_stride * calc_mi_size(cm->mi_rows)))
goto fail;
new_mi_size = cm->mi_stride * calc_mi_size(cm->mi_rows);
if (cm->mi_alloc_size < new_mi_size) {
cm->free_mi(cm);
if (cm->alloc_mi(cm, new_mi_size))
goto fail;
}
// Create the segmentation map structure and set to 0.
free_seg_map(cm);
if (alloc_seg_map(cm, cm->mi_rows * cm->mi_cols))
goto fail;
if (cm->seg_map_alloc_size < cm->mi_rows * cm->mi_cols) {
// Create the segmentation map structure and set to 0.
free_seg_map(cm);
if (alloc_seg_map(cm, cm->mi_rows * cm->mi_cols))
goto fail;
}
cm->above_context = (ENTROPY_CONTEXT *)vpx_calloc(
2 * mi_cols_aligned_to_sb(cm->mi_cols) * MAX_MB_PLANE,
sizeof(*cm->above_context));
if (!cm->above_context) goto fail;
if (cm->above_context_alloc_cols < cm->mi_cols) {
vpx_free(cm->above_context);
cm->above_context = (ENTROPY_CONTEXT *)vpx_calloc(
2 * mi_cols_aligned_to_sb(cm->mi_cols) * MAX_MB_PLANE,
sizeof(*cm->above_context));
if (!cm->above_context) goto fail;
cm->above_seg_context = (PARTITION_CONTEXT *)vpx_calloc(
mi_cols_aligned_to_sb(cm->mi_cols), sizeof(*cm->above_seg_context));
if (!cm->above_seg_context) goto fail;
vpx_free(cm->above_seg_context);
cm->above_seg_context = (PARTITION_CONTEXT *)vpx_calloc(
mi_cols_aligned_to_sb(cm->mi_cols), sizeof(*cm->above_seg_context));
if (!cm->above_seg_context) goto fail;
cm->above_context_alloc_cols = cm->mi_cols;
}
return 0;
@ -142,7 +158,6 @@ int vp9_alloc_context_buffers(VP9_COMMON *cm, int width, int height) {
}
void vp9_remove_common(VP9_COMMON *cm) {
vp9_free_ref_frame_buffers(cm);
vp9_free_context_buffers(cm);
vpx_free(cm->fc);
@ -154,7 +169,7 @@ void vp9_remove_common(VP9_COMMON *cm) {
void vp9_init_context_buffers(VP9_COMMON *cm) {
cm->setup_mi(cm);
if (cm->last_frame_seg_map && !cm->frame_parallel_decode)
vpx_memset(cm->last_frame_seg_map, 0, cm->mi_rows * cm->mi_cols);
memset(cm->last_frame_seg_map, 0, cm->mi_rows * cm->mi_cols);
}
void vp9_swap_current_and_last_seg_map(VP9_COMMON *cm) {

Некоторые файлы не были показаны из-за слишком большого количества измененных файлов Показать больше