FFmpeg
vp9dsp_init_16bpp.c
Go to the documentation of this file.
1 /*
2  * VP9 SIMD optimizations
3  *
4  * Copyright (c) 2013 Ronald S. Bultje <rsbultje gmail com>
5  *
6  * This file is part of FFmpeg.
7  *
8  * FFmpeg is free software; you can redistribute it and/or
9  * modify it under the terms of the GNU Lesser General Public
10  * License as published by the Free Software Foundation; either
11  * version 2.1 of the License, or (at your option) any later version.
12  *
13  * FFmpeg is distributed in the hope that it will be useful,
14  * but WITHOUT ANY WARRANTY; without even the implied warranty of
15  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
16  * Lesser General Public License for more details.
17  *
18  * You should have received a copy of the GNU Lesser General Public
19  * License along with FFmpeg; if not, write to the Free Software
20  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
21  */
22 
23 #include "libavutil/attributes.h"
24 #include "libavutil/cpu.h"
25 #include "libavutil/mem.h"
26 #include "libavutil/x86/cpu.h"
27 #include "libavcodec/vp9dsp.h"
29 
30 #if HAVE_X86ASM
31 
32 decl_fpel_func(put, 8, , mmx);
33 decl_fpel_func(avg, 8, _16, mmxext);
34 decl_fpel_func(put, 16, , sse);
35 decl_fpel_func(put, 32, , sse);
36 decl_fpel_func(put, 64, , sse);
37 decl_fpel_func(put, 128, , sse);
38 decl_fpel_func(avg, 16, _16, sse2);
39 decl_fpel_func(avg, 32, _16, sse2);
40 decl_fpel_func(avg, 64, _16, sse2);
41 decl_fpel_func(avg, 128, _16, sse2);
42 decl_fpel_func(put, 32, , avx);
43 decl_fpel_func(put, 64, , avx);
44 decl_fpel_func(put, 128, , avx);
45 decl_fpel_func(avg, 32, _16, avx2);
46 decl_fpel_func(avg, 64, _16, avx2);
47 decl_fpel_func(avg, 128, _16, avx2);
48 
49 decl_ipred_fns(v, 16, mmx, sse);
50 decl_ipred_fns(h, 16, mmxext, sse2);
51 decl_ipred_fns(dc, 16, mmxext, sse2);
52 decl_ipred_fns(dc_top, 16, mmxext, sse2);
53 decl_ipred_fns(dc_left, 16, mmxext, sse2);
54 decl_ipred_fn(dl, 16, 16, avx2);
55 decl_ipred_fn(dl, 32, 16, avx2);
56 decl_ipred_fn(dr, 16, 16, avx2);
57 decl_ipred_fn(dr, 32, 16, avx2);
58 
59 #define decl_ipred_dir_funcs(type) \
60 decl_ipred_fns(type, 16, sse2, sse2); \
61 decl_ipred_fns(type, 16, ssse3, ssse3); \
62 decl_ipred_fns(type, 16, avx, avx)
63 
64 decl_ipred_dir_funcs(dl);
65 decl_ipred_dir_funcs(dr);
66 decl_ipred_dir_funcs(vl);
67 decl_ipred_dir_funcs(vr);
68 decl_ipred_dir_funcs(hu);
69 decl_ipred_dir_funcs(hd);
70 #endif /* HAVE_X86ASM */
71 
73 {
74 #if HAVE_X86ASM
76 
77  if (EXTERNAL_MMX(cpu_flags)) {
78  init_fpel_func(4, 0, 8, put, , mmx);
79  init_ipred_func(v, VERT, 4, 16, mmx);
80  }
81 
83  init_fpel_func(4, 1, 8, avg, _16, mmxext);
84  init_ipred_func(h, HOR, 4, 16, mmxext);
85  init_ipred_func(dc, DC, 4, 16, mmxext);
86  init_ipred_func(dc_top, TOP_DC, 4, 16, mmxext);
87  init_ipred_func(dc_left, LEFT_DC, 4, 16, mmxext);
88  }
89 
90  if (EXTERNAL_SSE(cpu_flags)) {
91  init_fpel_func(3, 0, 16, put, , sse);
92  init_fpel_func(2, 0, 32, put, , sse);
93  init_fpel_func(1, 0, 64, put, , sse);
94  init_fpel_func(0, 0, 128, put, , sse);
95  init_8_16_32_ipred_funcs(v, VERT, 16, sse);
96  }
97 
98  if (EXTERNAL_SSE2(cpu_flags)) {
99  init_fpel_func(3, 1, 16, avg, _16, sse2);
100  init_fpel_func(2, 1, 32, avg, _16, sse2);
101  init_fpel_func(1, 1, 64, avg, _16, sse2);
102  init_fpel_func(0, 1, 128, avg, _16, sse2);
103  init_8_16_32_ipred_funcs(h, HOR, 16, sse2);
104  init_8_16_32_ipred_funcs(dc, DC, 16, sse2);
105  init_8_16_32_ipred_funcs(dc_top, TOP_DC, 16, sse2);
106  init_8_16_32_ipred_funcs(dc_left, LEFT_DC, 16, sse2);
107  init_ipred_funcs(dl, DIAG_DOWN_LEFT, 16, sse2);
108  init_ipred_funcs(dr, DIAG_DOWN_RIGHT, 16, sse2);
109  init_ipred_funcs(vl, VERT_LEFT, 16, sse2);
110  init_ipred_funcs(vr, VERT_RIGHT, 16, sse2);
111  init_ipred_funcs(hu, HOR_UP, 16, sse2);
112  init_ipred_funcs(hd, HOR_DOWN, 16, sse2);
113  }
114 
115  if (EXTERNAL_SSSE3(cpu_flags)) {
116  init_ipred_funcs(dl, DIAG_DOWN_LEFT, 16, ssse3);
117  init_ipred_funcs(dr, DIAG_DOWN_RIGHT, 16, ssse3);
118  init_ipred_funcs(vl, VERT_LEFT, 16, ssse3);
119  init_ipred_funcs(vr, VERT_RIGHT, 16, ssse3);
120  init_ipred_funcs(hu, HOR_UP, 16, ssse3);
121  init_ipred_funcs(hd, HOR_DOWN, 16, ssse3);
122  }
123 
125  init_fpel_func(2, 0, 32, put, , avx);
126  init_fpel_func(1, 0, 64, put, , avx);
127  init_fpel_func(0, 0, 128, put, , avx);
128  init_ipred_funcs(dl, DIAG_DOWN_LEFT, 16, avx);
129  init_ipred_funcs(dr, DIAG_DOWN_RIGHT, 16, avx);
130  init_ipred_funcs(vl, VERT_LEFT, 16, avx);
131  init_ipred_funcs(vr, VERT_RIGHT, 16, avx);
132  init_ipred_funcs(hu, HOR_UP, 16, avx);
133  init_ipred_funcs(hd, HOR_DOWN, 16, avx);
134  }
135 
137  init_fpel_func(2, 1, 32, avg, _16, avx2);
138  init_fpel_func(1, 1, 64, avg, _16, avx2);
139  init_fpel_func(0, 1, 128, avg, _16, avx2);
140  init_ipred_func(dl, DIAG_DOWN_LEFT, 16, 16, avx2);
141  init_ipred_func(dl, DIAG_DOWN_LEFT, 32, 16, avx2);
142  init_ipred_func(dr, DIAG_DOWN_RIGHT, 16, 16, avx2);
143 #if ARCH_X86_64
144  init_ipred_func(dr, DIAG_DOWN_RIGHT, 32, 16, avx2);
145 #endif
146  }
147 
148 #endif /* HAVE_X86ASM */
149 }
cpu.h
EXTERNAL_AVX2_FAST
#define EXTERNAL_AVX2_FAST(flags)
Definition: cpu.h:79
VP9DSPContext
Definition: vp9dsp.h:39
init_8_16_32_ipred_funcs
#define init_8_16_32_ipred_funcs(type, enum, bpp, opt)
Definition: vp9dsp_init.h:176
av_get_cpu_flags
int av_get_cpu_flags(void)
Return the flags which specify extensions supported by the CPU.
Definition: cpu.c:93
cpu_flags
static atomic_int cpu_flags
Definition: cpu.c:50
init_ipred_funcs
#define init_ipred_funcs(type, enum, bpp, opt)
Definition: vp9dsp_init.h:181
av_cold
#define av_cold
Definition: attributes.h:90
EXTERNAL_SSE
#define EXTERNAL_SSE(flags)
Definition: cpu.h:58
sse
static int sse(MpegEncContext *s, uint8_t *src1, uint8_t *src2, int w, int h, int stride)
Definition: mpegvideo_enc.c:2704
EXTERNAL_AVX_FAST
#define EXTERNAL_AVX_FAST(flags)
Definition: cpu.h:71
decl_ipred_fns
#define decl_ipred_fns(type, bpp, opt4, opt8_16_32)
Definition: vp9dsp_init.h:53
vp9dsp.h
dc
Tag MUST be and< 10hcoeff half pel interpolation filter coefficients, hcoeff[0] are the 2 middle coefficients[1] are the next outer ones and so on, resulting in a filter like:...eff[2], hcoeff[1], hcoeff[0], hcoeff[0], hcoeff[1], hcoeff[2] ... the sign of the coefficients is not explicitly stored but alternates after each coeff and coeff[0] is positive, so ...,+,-,+,-,+,+,-,+,-,+,... hcoeff[0] is not explicitly stored but found by subtracting the sum of all stored coefficients with signs from 32 hcoeff[0]=32 - hcoeff[1] - hcoeff[2] - ... a good choice for hcoeff and htaps is htaps=6 hcoeff={40,-10, 2} an alternative which requires more computations at both encoder and decoder side and may or may not be better is htaps=8 hcoeff={42,-14, 6,-2}ref_frames minimum of the number of available reference frames and max_ref_frames for example the first frame after a key frame always has ref_frames=1spatial_decomposition_type wavelet type 0 is a 9/7 symmetric compact integer wavelet 1 is a 5/3 symmetric compact integer wavelet others are reserved stored as delta from last, last is reset to 0 if always_reset||keyframeqlog quality(logarithmic quantizer scale) stored as delta from last, last is reset to 0 if always_reset||keyframemv_scale stored as delta from last, last is reset to 0 if always_reset||keyframe FIXME check that everything works fine if this changes between framesqbias dequantization bias stored as delta from last, last is reset to 0 if always_reset||keyframeblock_max_depth maximum depth of the block tree stored as delta from last, last is reset to 0 if always_reset||keyframequant_table quantization tableHighlevel bitstream structure:==============================--------------------------------------------|Header|--------------------------------------------|------------------------------------|||Block0||||split?||||yes no||||......... intra?||||:Block01 :yes no||||:Block02 :....... ..........||||:Block03 ::y DC ::ref index:||||:Block04 ::cb DC ::motion x :||||......... :cr DC ::motion y :||||....... ..........|||------------------------------------||------------------------------------|||Block1|||...|--------------------------------------------|------------ ------------ ------------|||Y subbands||Cb subbands||Cr subbands||||--- ---||--- ---||--- ---|||||LL0||HL0||||LL0||HL0||||LL0||HL0|||||--- ---||--- ---||--- ---||||--- ---||--- ---||--- ---|||||LH0||HH0||||LH0||HH0||||LH0||HH0|||||--- ---||--- ---||--- ---||||--- ---||--- ---||--- ---|||||HL1||LH1||||HL1||LH1||||HL1||LH1|||||--- ---||--- ---||--- ---||||--- ---||--- ---||--- ---|||||HH1||HL2||||HH1||HL2||||HH1||HL2|||||...||...||...|||------------ ------------ ------------|--------------------------------------------Decoding process:=================------------|||Subbands|------------||||------------|Intra DC||||LL0 subband prediction ------------|\ Dequantization ------------------- \||Reference frames|\ IDWT|------- -------|Motion \|||Frame 0||Frame 1||Compensation . OBMC v -------|------- -------|--------------. \------> Frame n output Frame Frame<----------------------------------/|...|------------------- Range Coder:============Binary Range Coder:------------------- The implemented range coder is an adapted version based upon "Range encoding: an algorithm for removing redundancy from a digitised message." by G. N. N. Martin. The symbols encoded by the Snow range coder are bits(0|1). The associated probabilities are not fix but change depending on the symbol mix seen so far. bit seen|new state ---------+----------------------------------------------- 0|256 - state_transition_table[256 - old_state];1|state_transition_table[old_state];state_transition_table={ 0, 0, 0, 0, 0, 0, 0, 0, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 94, 95, 96, 97, 98, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 133, 134, 135, 136, 137, 138, 139, 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, 152, 152, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 163, 164, 165, 166, 167, 168, 169, 170, 171, 171, 172, 173, 174, 175, 176, 177, 178, 179, 180, 181, 182, 183, 184, 185, 186, 187, 188, 189, 190, 190, 191, 192, 194, 194, 195, 196, 197, 198, 199, 200, 201, 202, 202, 204, 205, 206, 207, 208, 209, 209, 210, 211, 212, 213, 215, 215, 216, 217, 218, 219, 220, 220, 222, 223, 224, 225, 226, 227, 227, 229, 229, 230, 231, 232, 234, 234, 235, 236, 237, 238, 239, 240, 241, 242, 243, 244, 245, 246, 247, 248, 248, 0, 0, 0, 0, 0, 0, 0};FIXME Range Coding of integers:------------------------- FIXME Neighboring Blocks:===================left and top are set to the respective blocks unless they are outside of the image in which case they are set to the Null block top-left is set to the top left block unless it is outside of the image in which case it is set to the left block if this block has no larger parent block or it is at the left side of its parent block and the top right block is not outside of the image then the top right block is used for top-right else the top-left block is used Null block y, cb, cr are 128 level, ref, mx and my are 0 Motion Vector Prediction:=========================1. the motion vectors of all the neighboring blocks are scaled to compensate for the difference of reference frames scaled_mv=(mv *(256 *(current_reference+1)/(mv.reference+1))+128)> the median of the scaled top and top right vectors is used as motion vector prediction the used motion vector is the sum of the predictor and(mvx_diff, mvy_diff) *mv_scale Intra DC Prediction block[y][x] dc[1]
Definition: snow.txt:400
cpu.h
avg
#define avg(a, b, c, d)
Definition: colorspacedsp_template.c:28
attributes.h
EXTERNAL_SSE2
#define EXTERNAL_SSE2(flags)
Definition: cpu.h:59
decl_fpel_func
#define decl_fpel_func(avg, sz, bpp, opt)
Definition: vp9dsp_init.h:31
init_ipred_func
#define init_ipred_func(type, enum, sz, bpp, opt)
Definition: vp9dsp_init.h:172
decl_ipred_fn
#define decl_ipred_fn(type, sz, bpp, opt)
Definition: vp9dsp_init.h:47
ff_vp9dsp_init_16bpp_x86
av_cold void ff_vp9dsp_init_16bpp_x86(VP9DSPContext *dsp)
Definition: vp9dsp_init_16bpp.c:72
init_fpel_func
#define init_fpel_func(idx1, idx2, sz, type, bpp, opt)
Definition: vp9dsp_init.h:140
mem.h
vp9dsp_init.h
h
h
Definition: vp9dsp_template.c:2038
EXTERNAL_SSSE3
#define EXTERNAL_SSSE3(flags)
Definition: cpu.h:65
EXTERNAL_MMX
#define EXTERNAL_MMX(flags)
Definition: cpu.h:56
EXTERNAL_MMXEXT
#define EXTERNAL_MMXEXT(flags)
Definition: cpu.h:57