libavutil/fixed_dsp.c

   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 (nedeljko.babic imgtec 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 #include "common.h"
  49 #include "fixed_dsp.h"
  50 #include "mem.h"
  51
  52 static void vector_fmul_add_c(int *dst, const int *src0, const int *src1, const int *src2, int len){
  53     int i;
  54     int64_t accu;
  55
  56     for (i=0; i<len; i++) {
  57         accu = (int64_t)src0[i] * src1[i];
  58         dst[i] = src2[i] + (int)((accu + 0x40000000) >> 31);
  59     }
  60 }
  61
  62 static void vector_fmul_reverse_c(int *dst, const int *src0, const int *src1, int len)
  63 {
  64     int i;
  65     int64_t accu;
  66
  67     src1 += len-1;
  68     for (i=0; i<len; i++) {
  69         accu = (int64_t)src0[i] * src1[-i];
  70         dst[i] = (int)((accu+0x40000000) >> 31);
  71     }
  72 }
  73
  74 static void vector_fmul_window_scaled_c(int16_t *dst, const int32_t *src0,
  75                                        const int32_t *src1, const int32_t *win,
  76                                        int len, uint8_t bits)
  77 {
  78     int32_t s0, s1, wi, wj, i,j, round;
  79
  80     dst += len;
  81     win += len;
  82     src0+= len;
  83     round = bits? 1 << (bits-1) : 0;
  84
  85     for (i=-len, j=len-1; i<0; i++, j--) {
  86         s0 = src0[i];
  87         s1 = src1[j];
  88         wi = win[i];
  89         wj = win[j];
  90         dst[i] = av_clip_int16(((((int64_t)s0*wj - (int64_t)s1*wi + 0x40000000) >> 31) + round) >> bits);
  91         dst[j] = av_clip_int16(((((int64_t)s0*wi + (int64_t)s1*wj + 0x40000000) >> 31) + round) >> bits);
  92     }
  93 }
  94
  95 static void vector_fmul_window_c(int32_t *dst, const int32_t *src0,
  96                                        const int32_t *src1, const int32_t *win,
  97                                        int len)
  98 {
  99     int32_t s0, s1, wi, wj, i, j;
 100
 101     dst += len;
 102     win += len;
 103     src0+= len;
 104
 105     for (i=-len, j=len-1; i<0; i++, j--) {
 106         s0 = src0[i];
 107         s1 = src1[j];
 108         wi = win[i];
 109         wj = win[j];
 110         dst[i] = ((int64_t)s0*wj - (int64_t)s1*wi + 0x40000000) >> 31;
 111         dst[j] = ((int64_t)s0*wi + (int64_t)s1*wj + 0x40000000) >> 31;
 112     }
 113 }
 114
 115 static void vector_fmul_c(int *dst, const int *src0, const int *src1, int len)
 116 {
 117     int i;
 118     int64_t accu;
 119
 120     for (i = 0; i < len; i++){
 121         accu = (int64_t)src0[i] * src1[i];
 122         dst[i] = (int)((accu+0x40000000) >> 31);
 123     }
 124 }
 125
 126 static int scalarproduct_fixed_c(const int *v1, const int *v2, int len)
 127 {
 128     /** p is initialized with 0x40000000 so that the proper rounding will occur
 129       * at the end */
 130     int64_t p = 0x40000000;
 131     int i;
 132
 133     for (i = 0; i < len; i++)
 134         p += (int64_t)v1[i] * v2[i];
 135
 136     return (int)(p >> 31);
 137 }
 138
 139 static void butterflies_fixed_c(int *restrict v1s, int *restrict v2, int len)
 140 {
 141     int i;
 142     unsigned int *v1 = v1s;
 143
 144     for (i = 0; i < len; i++){
 145         int t = v1[i] - v2[i];
 146         v1[i] += v2[i];
 147         v2[i] = t;
 148     }
 149 }
 150
 151 AVFixedDSPContext * avpriv_alloc_fixed_dsp(int bit_exact)
 152 {
 153     AVFixedDSPContext * fdsp = av_malloc(sizeof(AVFixedDSPContext));
 154
 155     if (!fdsp)
 156         return NULL;
 157
 158     fdsp->vector_fmul_window_scaled = vector_fmul_window_scaled_c;
 159     fdsp->vector_fmul_window = vector_fmul_window_c;
 160     fdsp->vector_fmul = vector_fmul_c;
 161     fdsp->vector_fmul_add = vector_fmul_add_c;
 162     fdsp->vector_fmul_reverse = vector_fmul_reverse_c;
 163     fdsp->butterflies_fixed = butterflies_fixed_c;
 164     fdsp->scalarproduct_fixed = scalarproduct_fixed_c;
 165
 166 #if ARCH_RISCV
 167     ff_fixed_dsp_init_riscv(fdsp);
 168 #elif ARCH_X86
 169     ff_fixed_dsp_init_x86(fdsp);
 170 #endif
 171
 172     return fdsp;
 173 }