summaryrefslogtreecommitdiffstats
path: root/libavutil/fixed_dsp.h
blob: 03987adddc5bedd6c476d2698a4c05c5bde66551 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
/*
 * Copyright (c) 2012
 *      MIPS Technologies, Inc., California.
 *
 * Redistribution and use in source and binary forms, with or without
 * modification, are permitted provided that the following conditions
 * are met:
 * 1. Redistributions of source code must retain the above copyright
 *    notice, this list of conditions and the following disclaimer.
 * 2. Redistributions in binary form must reproduce the above copyright
 *    notice, this list of conditions and the following disclaimer in the
 *    documentation and/or other materials provided with the distribution.
 * 3. Neither the name of the MIPS Technologies, Inc., nor the names of its
 *    contributors may be used to endorse or promote products derived from
 *    this software without specific prior written permission.
 *
 * THIS SOFTWARE IS PROVIDED BY THE MIPS TECHNOLOGIES, INC. ``AS IS'' AND
 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
 * ARE DISCLAIMED.  IN NO EVENT SHALL THE MIPS TECHNOLOGIES, INC. BE LIABLE
 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
 * SUCH DAMAGE.
 *
 * Author:  Nedeljko Babic (nbabic@mips.com)
 *
 * This file is part of FFmpeg.
 *
 * FFmpeg is free software; you can redistribute it and/or
 * modify it under the terms of the GNU Lesser General Public
 * License as published by the Free Software Foundation; either
 * version 2.1 of the License, or (at your option) any later version.
 *
 * FFmpeg is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
 * Lesser General Public License for more details.
 *
 * You should have received a copy of the GNU Lesser General Public
 * License along with FFmpeg; if not, write to the Free Software
 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
 */

#ifndef AVUTIL_FIXED_DSP_H
#define AVUTIL_FIXED_DSP_H

#include <stdint.h>
#include "attributes.h"
#include "common.h"
#include "libavcodec/mathops.h"

typedef struct AVFixedDSPContext {
    /* Assume len is a multiple of 16, and arrays are 32-byte aligned */
    /* Results of multiplications are scaled down by 31 bit (and rounded) if not
     * stated otherwise */

    /**
     * Overlap/add with window function.
     * Result is scaled down by "bits" bits.
     * Used primarily by MDCT-based audio codecs.
     * Source and destination vectors must overlap exactly or not at all.
     *
     * @param dst  result vector
     *             constraints: 16-byte aligned
     * @param src0 first source vector
     *             constraints: 16-byte aligned
     * @param src1 second source vector
     *             constraints: 16-byte aligned
     * @param win  half-window vector
     *             constraints: 16-byte aligned
     * @param len  length of vector
     *             constraints: multiple of 4
     * @param bits scaling parameter
     *
     */
    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.
     * Used primarily by MDCT-based audio codecs.
     * Source and destination vectors must overlap exactly or not at all.
     *
     * @param dst  result vector
     *             constraints: 32-byte aligned
     * @param src0 first source vector
     *             constraints: 16-byte aligned
     * @param src1 second source vector
     *             constraints: 16-byte aligned
     * @param win  half-window vector
     *             constraints: 16-byte aligned
     * @param len  length of vector
     *             constraints: multiple of 4
     */
    void (*vector_fmul_window)(int32_t *dst, const int32_t *src0, const int32_t *src1, const int32_t *win, int len);

    /**
     * Fixed-point multiplication that calculates the entry wise product of two
     * vectors of integers and stores the result in a vector of integers.
     *
     * @param dst  output vector
     *             constraints: 32-byte aligned
     * @param src0 first input vector
     *             constraints: 32-byte aligned
     * @param src1 second input vector
     *             constraints: 32-byte aligned
     * @param len  number of elements in the input
     *             constraints: multiple of 16
     */
    void (*vector_fmul)(int *dst, const int *src0, const int *src1,
                        int len);

    void (*vector_fmul_reverse)(int *dst, const int *src0, const int *src1, int len);
    /**
     * Calculate the entry wise product of two vectors of integers, add a third vector of
     * integers and store the result in a vector of integers.
     *
     * @param dst  output vector
     *             constraints: 32-byte aligned
     * @param src0 first input vector
     *             constraints: 32-byte aligned
     * @param src1 second input vector
     *             constraints: 32-byte aligned
     * @param src2 third input vector
     *             constraints: 32-byte aligned
     * @param len  number of elements in the input
     *             constraints: multiple of 16
     */
    void (*vector_fmul_add)(int *dst, const int *src0, const int *src1,
                            const int *src2, int len);

    /**
     * Calculate the scalar product of two vectors of integers.
     *
     * @param v1  first vector, 16-byte aligned
     * @param v2  second vector, 16-byte aligned
     * @param len length of vectors, multiple of 4
     *
     * @return sum of elementwise products
     */
    int (*scalarproduct_fixed)(const int *v1, const int *v2, int len);

    /**
     * Calculate the sum and difference of two vectors of integers.
     *
     * @param v1  first input vector, sum output, 16-byte aligned
     * @param v2  second input vector, difference output, 16-byte aligned
     * @param len length of vectors, multiple of 4
     */
    void (*butterflies_fixed)(int *av_restrict v1, int *av_restrict v2, int len);
} AVFixedDSPContext;

/**
 * Allocate and initialize a fixed DSP context.
 * note: should be freed with a av_free call when no longer needed.
 *
 * @param strict  setting to non-zero avoids using functions which may not be IEEE-754 compliant
 */
AVFixedDSPContext * avpriv_alloc_fixed_dsp(int strict);

/**
 * Calculate the square root
 *
 * @param x    input fixed point number
 *
 * @param bits format of fixed point number (32 - bits).bits
 *
 * note: input is normalized to (0, 1) fixed point value
 */

static av_always_inline int fixed_sqrt(int x, int bits)
{
    int retval, bit_mask, guess, square, i;
    int64_t accu;
    int shift1 = 30 - bits;
    int shift2 = bits - 15;

    if (shift1 > 0) retval = ff_sqrt(x << shift1);
    else retval = ff_sqrt(x >> -shift1);

    if (shift2 > 0) {
        retval = retval << shift2;
        bit_mask = (1 << (shift2 - 1));

        for (i=0; i<shift2; i++){
            guess = retval + bit_mask;
            accu = (int64_t)guess * guess;
            square = (int)((accu + bit_mask) >> bits);
            if (x >= square)
                retval += bit_mask;
            bit_mask >>= 1;
        }

    }
    else retval >>= (-shift2);

    return retval;
}

#endif /* AVUTIL_FIXED_DSP_H */
OpenPOWER on IntegriCloud