FFmpeg  4.3
fixed_dsp.h
Go to the documentation of this file.
1 /*
2  * Copyright (c) 2012
3  * MIPS Technologies, Inc., California.
4  *
5  * Redistribution and use in source and binary forms, with or without
6  * modification, are permitted provided that the following conditions
7  * are met:
8  * 1. Redistributions of source code must retain the above copyright
9  * notice, this list of conditions and the following disclaimer.
10  * 2. Redistributions in binary form must reproduce the above copyright
11  * notice, this list of conditions and the following disclaimer in the
12  * documentation and/or other materials provided with the distribution.
13  * 3. Neither the name of the MIPS Technologies, Inc., nor the names of its
14  * contributors may be used to endorse or promote products derived from
15  * this software without specific prior written permission.
16  *
17  * THIS SOFTWARE IS PROVIDED BY THE MIPS TECHNOLOGIES, INC. ``AS IS'' AND
18  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
19  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
20  * ARE DISCLAIMED. IN NO EVENT SHALL THE MIPS TECHNOLOGIES, INC. BE LIABLE
21  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
22  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
23  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
24  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
25  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
26  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
27  * SUCH DAMAGE.
28  *
29  * Author: Nedeljko Babic (nbabic@mips.com)
30  *
31  * This file is part of FFmpeg.
32  *
33  * FFmpeg is free software; you can redistribute it and/or
34  * modify it under the terms of the GNU Lesser General Public
35  * License as published by the Free Software Foundation; either
36  * version 2.1 of the License, or (at your option) any later version.
37  *
38  * FFmpeg is distributed in the hope that it will be useful,
39  * but WITHOUT ANY WARRANTY; without even the implied warranty of
40  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
41  * Lesser General Public License for more details.
42  *
43  * You should have received a copy of the GNU Lesser General Public
44  * License along with FFmpeg; if not, write to the Free Software
45  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
46  */
47 
48 #ifndef AVUTIL_FIXED_DSP_H
49 #define AVUTIL_FIXED_DSP_H
50 
51 #include <stdint.h>
52 #include "attributes.h"
53 #include "common.h"
54 #include "libavcodec/mathops.h"
55 
56 typedef struct AVFixedDSPContext {
57  /* Assume len is a multiple of 16, and arrays are 32-byte aligned */
58  /* Results of multiplications are scaled down by 31 bit (and rounded) if not
59  * stated otherwise */
60 
61  /**
62  * Overlap/add with window function.
63  * Result is scaled down by "bits" bits.
64  * Used primarily by MDCT-based audio codecs.
65  * Source and destination vectors must overlap exactly or not at all.
66  *
67  * @param dst result vector
68  * constraints: 16-byte aligned
69  * @param src0 first source vector
70  * constraints: 16-byte aligned
71  * @param src1 second source vector
72  * constraints: 16-byte aligned
73  * @param win half-window vector
74  * constraints: 16-byte aligned
75  * @param len length of vector
76  * constraints: multiple of 4
77  * @param bits scaling parameter
78  *
79  */
80  void (*vector_fmul_window_scaled)(int16_t *dst, const int32_t *src0, const int32_t *src1, const int32_t *win, int len, uint8_t bits);
81 
82  /**
83  * Overlap/add with window function.
84  * Used primarily by MDCT-based audio codecs.
85  * Source and destination vectors must overlap exactly or not at all.
86  *
87  * @param dst result vector
88  * constraints: 32-byte aligned
89  * @param src0 first source vector
90  * constraints: 16-byte aligned
91  * @param src1 second source vector
92  * constraints: 16-byte aligned
93  * @param win half-window vector
94  * constraints: 16-byte aligned
95  * @param len length of vector
96  * constraints: multiple of 4
97  */
98  void (*vector_fmul_window)(int32_t *dst, const int32_t *src0, const int32_t *src1, const int32_t *win, int len);
99 
100  /**
101  * Fixed-point multiplication that calculates the entry wise product of two
102  * vectors of integers and stores the result in a vector of integers.
103  *
104  * @param dst output vector
105  * constraints: 32-byte aligned
106  * @param src0 first input vector
107  * constraints: 32-byte aligned
108  * @param src1 second input vector
109  * constraints: 32-byte aligned
110  * @param len number of elements in the input
111  * constraints: multiple of 16
112  */
113  void (*vector_fmul)(int *dst, const int *src0, const int *src1,
114  int len);
115 
116  void (*vector_fmul_reverse)(int *dst, const int *src0, const int *src1, int len);
117  /**
118  * Calculate the entry wise product of two vectors of integers, add a third vector of
119  * integers and store the result in a vector of integers.
120  *
121  * @param dst output vector
122  * constraints: 32-byte aligned
123  * @param src0 first input vector
124  * constraints: 32-byte aligned
125  * @param src1 second input vector
126  * constraints: 32-byte aligned
127  * @param src2 third input vector
128  * constraints: 32-byte aligned
129  * @param len number of elements in the input
130  * constraints: multiple of 16
131  */
132  void (*vector_fmul_add)(int *dst, const int *src0, const int *src1,
133  const int *src2, int len);
134 
135  /**
136  * Calculate the scalar product of two vectors of integers.
137  *
138  * @param v1 first vector, 16-byte aligned
139  * @param v2 second vector, 16-byte aligned
140  * @param len length of vectors, multiple of 4
141  *
142  * @return sum of elementwise products
143  */
144  int (*scalarproduct_fixed)(const int *v1, const int *v2, int len);
145 
146  /**
147  * Calculate the sum and difference of two vectors of integers.
148  *
149  * @param v1 first input vector, sum output, 16-byte aligned
150  * @param v2 second input vector, difference output, 16-byte aligned
151  * @param len length of vectors, multiple of 4
152  */
153  void (*butterflies_fixed)(int *av_restrict v1, int *av_restrict v2, int len);
155 
156 /**
157  * Allocate and initialize a fixed DSP context.
158  * note: should be freed with a av_free call when no longer needed.
159  *
160  * @param strict setting to non-zero avoids using functions which may not be IEEE-754 compliant
161  */
163 
165 
166 /**
167  * Calculate the square root
168  *
169  * @param x input fixed point number
170  *
171  * @param bits format of fixed point number (32 - bits).bits
172  *
173  * note: input is normalized to (0, 1) fixed point value
174  */
175 
176 static av_always_inline int fixed_sqrt(int x, int bits)
177 {
178  int retval, bit_mask, guess, square, i;
179  int64_t accu;
180  int shift1 = 30 - bits;
181  int shift2 = bits - 15;
182 
183  if (shift1 > 0) retval = ff_sqrt(x << shift1);
184  else retval = ff_sqrt(x >> -shift1);
185 
186  if (shift2 > 0) {
187  retval = retval << shift2;
188  bit_mask = (1 << (shift2 - 1));
189 
190  for (i=0; i<shift2; i++){
191  guess = retval + bit_mask;
192  accu = (int64_t)guess * guess;
193  square = (int)((accu + bit_mask) >> bits);
194  if (x >= square)
195  retval += bit_mask;
196  bit_mask >>= 1;
197  }
198 
199  }
200  else retval >>= (-shift2);
201 
202  return retval;
203 }
204 
205 #endif /* AVUTIL_FIXED_DSP_H */
av_restrict
#define av_restrict
Definition: config.h:10
AVFixedDSPContext
Definition: fixed_dsp.h:56
win
static float win(SuperEqualizerContext *s, float n, int N)
Definition: af_superequalizer.c:119
AVFixedDSPContext::vector_fmul_add
void(* vector_fmul_add)(int *dst, const int *src0, const int *src1, const int *src2, int len)
Calculate the entry wise product of two vectors of integers, add a third vector of integers and store...
Definition: fixed_dsp.h:132
x
FFmpeg Automated Testing Environment ************************************Introduction Using FATE from your FFmpeg source directory Submitting the results to the FFmpeg result aggregation server Uploading new samples to the fate suite FATE makefile targets and variables Makefile targets Makefile variables Examples Introduction **************FATE is an extended regression suite on the client side and a means for results aggregation and presentation on the server side The first part of this document explains how you can use FATE from your FFmpeg source directory to test your ffmpeg binary The second part describes how you can run FATE to submit the results to FFmpeg’s FATE server In any way you can have a look at the publicly viewable FATE results by visiting this as it can be seen if some test on some platform broke with their recent contribution This usually happens on the platforms the developers could not test on The second part of this document describes how you can run FATE to submit your results to FFmpeg’s FATE server If you want to submit your results be sure to check that your combination of OS and compiler is not already listed on the above mentioned website In the third part you can find a comprehensive listing of FATE makefile targets and variables Using FATE from your FFmpeg source directory **********************************************If you want to run FATE on your machine you need to have the samples in place You can get the samples via the build target fate rsync Use this command from the top level source this will cause FATE to fail NOTE To use a custom wrapper to run the pass ‘ target exec’ to ‘configure’ or set the TARGET_EXEC Make variable Submitting the results to the FFmpeg result aggregation server ****************************************************************To submit your results to the server you should run fate through the shell script ‘tests fate sh’ from the FFmpeg sources This script needs to be invoked with a configuration file as its first argument tests fate sh path to fate_config A configuration file template with comments describing the individual configuration variables can be found at ‘doc fate_config sh template’ Create a configuration that suits your based on the configuration template The ‘slot’ configuration variable can be any string that is not yet but it is suggested that you name it adhering to the following pattern ‘ARCH OS COMPILER COMPILER VERSION’ The configuration file itself will be sourced in a shell therefore all shell features may be used This enables you to setup the environment as you need it for your build For your first test runs the ‘fate_recv’ variable should be empty or commented out This will run everything as normal except that it will omit the submission of the results to the server The following files should be present in $workdir as specified in the configuration it may help to try out the ‘ssh’ command with one or more ‘ v’ options You should get detailed output concerning your SSH configuration and the authentication process The only thing left is to automate the execution of the fate sh script and the synchronisation of the samples directory Uploading new samples to the fate suite *****************************************If you need a sample uploaded send a mail to samples request This is for developers who have an account on the fate suite server If you upload new please make sure they are as small as space on each network bandwidth and so on benefit from smaller test cases Also keep in mind older checkouts use existing sample that means in practice generally do not remove or overwrite files as it likely would break older checkouts or releases Also all needed samples for a commit should be ideally before the push If you need an account for frequently uploading samples or you wish to help others by doing that send a mail to ffmpeg devel rsync vauL Duo x
Definition: fate.txt:150
ff_sqrt
#define ff_sqrt
Definition: mathops.h:206
bits
uint8_t bits
Definition: vp3data.h:202
AVFixedDSPContext::scalarproduct_fixed
int(* scalarproduct_fixed)(const int *v1, const int *v2, int len)
Calculate the scalar product of two vectors of integers.
Definition: fixed_dsp.h:144
int32_t
int32_t
Definition: audio_convert.c:194
mathops.h
AVFixedDSPContext::vector_fmul_reverse
void(* vector_fmul_reverse)(int *dst, const int *src0, const int *src1, int len)
Definition: fixed_dsp.h:116
shift1
static const int shift1[6]
Definition: dxa.c:50
AVFixedDSPContext::vector_fmul
void(* vector_fmul)(int *dst, const int *src0, const int *src1, int len)
Fixed-point multiplication that calculates the entry wise product of two vectors of integers and stor...
Definition: fixed_dsp.h:113
attributes.h
shift2
static const int shift2[6]
Definition: dxa.c:51
src0
#define src0
Definition: h264pred.c:138
AVFixedDSPContext::butterflies_fixed
void(* butterflies_fixed)(int *av_restrict v1, int *av_restrict v2, int len)
Calculate the sum and difference of two vectors of integers.
Definition: fixed_dsp.h:153
src1
#define src1
Definition: h264pred.c:139
i
#define i(width, name, range_min, range_max)
Definition: cbs_h2645.c:269
common.h
av_always_inline
#define av_always_inline
Definition: attributes.h:49
uint8_t
uint8_t
Definition: audio_convert.c:194
len
int len
Definition: vorbis_enc_data.h:452
square
static int square(int x)
Definition: roqvideoenc.c:113
void
typedef void(RENAME(mix_any_func_type))
Definition: rematrix_template.c:52
ff_fixed_dsp_init_x86
void ff_fixed_dsp_init_x86(AVFixedDSPContext *fdsp)
Definition: fixed_dsp_init.c:28
AVFixedDSPContext::vector_fmul_window_scaled
void(* vector_fmul_window_scaled)(int16_t *dst, const int32_t *src0, const int32_t *src1, const int32_t *win, int len, uint8_t bits)
Overlap/add with window function.
Definition: fixed_dsp.h:80
fixed_sqrt
static av_always_inline int fixed_sqrt(int x, int bits)
Calculate the square root.
Definition: fixed_dsp.h:176
avpriv_alloc_fixed_dsp
AVFixedDSPContext * avpriv_alloc_fixed_dsp(int strict)
Allocate and initialize a fixed DSP context.
Definition: fixed_dsp.c:148
int
int
Definition: ffmpeg_filter.c:192
AVFixedDSPContext::vector_fmul_window
void(* vector_fmul_window)(int32_t *dst, const int32_t *src0, const int32_t *src1, const int32_t *win, int len)
Overlap/add with window function.
Definition: fixed_dsp.h:98