ffmpeg/libavcodec/mathops.h

   1 /*
   2  * simple math operations
   3  * Copyright (c) 2001, 2002 Fabrice Bellard
   4  * Copyright (c) 2006 Michael Niedermayer <michaelni@gmx.at> et al
   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 #ifndef AVCODEC_MATHOPS_H
  23 #define AVCODEC_MATHOPS_H
  24
  25 #include <stdint.h>
  26
  27 #include "libavutil/common.h"
  28 #include "config.h"
  29
  30 #define MAX_NEG_CROP 1024
  31
  32 extern const uint32_t ff_inverse[257];
  33 extern const uint8_t  ff_reverse[256];
  34 extern const uint8_t ff_sqrt_tab[256];
  35 extern const uint8_t ff_crop_tab[256 + 2 * MAX_NEG_CROP];
  36 extern const uint8_t ff_zigzag_direct[64];
  37
  38 #if   ARCH_ARM
  39 #   include "arm/mathops.h"
  40 #elif ARCH_AVR32
  41 #   include "avr32/mathops.h"
  42 #elif ARCH_MIPS
  43 #   include "mips/mathops.h"
  44 #elif ARCH_PPC
  45 #   include "ppc/mathops.h"
  46 #elif ARCH_X86
  47 #   include "x86/mathops.h"
  48 #endif
  49
  50 /* generic implementation */
  51
  52 #ifndef MUL64
  53 #   define MUL64(a,b) ((int64_t)(a) * (int64_t)(b))
  54 #endif
  55
  56 #ifndef MULL
  57 #   define MULL(a,b,s) (MUL64(a, b) >> (s))
  58 #endif
  59
  60 #ifndef MULH
  61 static av_always_inline int MULH(int a, int b){
  62     return MUL64(a, b) >> 32;
  63 }
  64 #endif
  65
  66 #ifndef UMULH
  67 static av_always_inline unsigned UMULH(unsigned a, unsigned b){
  68     return ((uint64_t)(a) * (uint64_t)(b))>>32;
  69 }
  70 #endif
  71
  72 #ifndef MAC64
  73 #   define MAC64(d, a, b) ((d) += MUL64(a, b))
  74 #endif
  75
  76 #ifndef MLS64
  77 #   define MLS64(d, a, b) ((d) -= MUL64(a, b))
  78 #endif
  79
  80 /* signed 16x16 -> 32 multiply add accumulate */
  81 #ifndef MAC16
  82 #   define MAC16(rt, ra, rb) rt += (ra) * (rb)
  83 #endif
  84
  85 /* signed 16x16 -> 32 multiply */
  86 #ifndef MUL16
  87 #   define MUL16(ra, rb) ((ra) * (rb))
  88 #endif
  89
  90 #ifndef MLS16
  91 #   define MLS16(rt, ra, rb) ((rt) -= (ra) * (rb))
  92 #endif
  93
  94 /* median of 3 */
  95 #ifndef mid_pred
  96 #define mid_pred mid_pred
  97 static inline av_const int mid_pred(int a, int b, int c)
  98 {
  99 #if 0
 100     int t= (a-b)&((a-b)>>31);
 101     a-=t;
 102     b+=t;
 103     b-= (b-c)&((b-c)>>31);
 104     b+= (a-b)&((a-b)>>31);
 105
 106     return b;
 107 #else
 108     if(a>b){
 109         if(c>b){
 110             if(c>a) b=a;
 111             else    b=c;
 112         }
 113     }else{
 114         if(b>c){
 115             if(c>a) b=c;
 116             else    b=a;
 117         }
 118     }
 119     return b;
 120 #endif
 121 }
 122 #endif
 123
 124 #ifndef sign_extend
 125 static inline av_const int sign_extend(int val, unsigned bits)
 126 {
 127     unsigned shift = 8 * sizeof(int) - bits;
 128     union { unsigned u; int s; } v = { (unsigned) val << shift };
 129     return v.s >> shift;
 130 }
 131 #endif
 132
 133 #ifndef zero_extend
 134 static inline av_const unsigned zero_extend(unsigned val, unsigned bits)
 135 {
 136     return (val << ((8 * sizeof(int)) - bits)) >> ((8 * sizeof(int)) - bits);
 137 }
 138 #endif
 139
 140 #ifndef COPY3_IF_LT
 141 #define COPY3_IF_LT(x, y, a, b, c, d)\
 142 if ((y) < (x)) {\
 143     (x) = (y);\
 144     (a) = (b);\
 145     (c) = (d);\
 146 }
 147 #endif
 148
 149 #ifndef MASK_ABS
 150 #define MASK_ABS(mask, level) do {              \
 151         mask  = level >> 31;                    \
 152         level = (level ^ mask) - mask;          \
 153     } while (0)
 154 #endif
 155
 156 #ifndef NEG_SSR32
 157 #   define NEG_SSR32(a,s) ((( int32_t)(a))>>(32-(s)))
 158 #endif
 159
 160 #ifndef NEG_USR32
 161 #   define NEG_USR32(a,s) (((uint32_t)(a))>>(32-(s)))
 162 #endif
 163
 164 #if HAVE_BIGENDIAN
 165 # ifndef PACK_2U8
 166 #   define PACK_2U8(a,b)     (((a) <<  8) | (b))
 167 # endif
 168 # ifndef PACK_4U8
 169 #   define PACK_4U8(a,b,c,d) (((a) << 24) | ((b) << 16) | ((c) << 8) | (d))
 170 # endif
 171 # ifndef PACK_2U16
 172 #   define PACK_2U16(a,b)    (((a) << 16) | (b))
 173 # endif
 174 #else
 175 # ifndef PACK_2U8
 176 #   define PACK_2U8(a,b)     (((b) <<  8) | (a))
 177 # endif
 178 # ifndef PACK_4U2
 179 #   define PACK_4U8(a,b,c,d) (((d) << 24) | ((c) << 16) | ((b) << 8) | (a))
 180 # endif
 181 # ifndef PACK_2U16
 182 #   define PACK_2U16(a,b)    (((b) << 16) | (a))
 183 # endif
 184 #endif
 185
 186 #ifndef PACK_2S8
 187 #   define PACK_2S8(a,b)     PACK_2U8((a)&255, (b)&255)
 188 #endif
 189 #ifndef PACK_4S8
 190 #   define PACK_4S8(a,b,c,d) PACK_4U8((a)&255, (b)&255, (c)&255, (d)&255)
 191 #endif
 192 #ifndef PACK_2S16
 193 #   define PACK_2S16(a,b)    PACK_2U16((a)&0xffff, (b)&0xffff)
 194 #endif
 195
 196 #ifndef FASTDIV
 197 #   define FASTDIV(a,b) ((uint32_t)((((uint64_t)a) * ff_inverse[b]) >> 32))
 198 #endif /* FASTDIV */
 199
 200 static inline av_const unsigned int ff_sqrt(unsigned int a)
 201 {
 202     unsigned int b;
 203
 204     if (a < 255) return (ff_sqrt_tab[a + 1] - 1) >> 4;
 205     else if (a < (1 << 12)) b = ff_sqrt_tab[a >> 4] >> 2;
 206 #if !CONFIG_SMALL
 207     else if (a < (1 << 14)) b = ff_sqrt_tab[a >> 6] >> 1;
 208     else if (a < (1 << 16)) b = ff_sqrt_tab[a >> 8]   ;
 209 #endif
 210     else {
 211         int s = av_log2_16bit(a >> 16) >> 1;
 212         unsigned int c = a >> (s + 2);
 213         b = ff_sqrt_tab[c >> (s + 8)];
 214         b = FASTDIV(c,b) + (b << s);
 215     }
 216
 217     return b - (a < b * b);
 218 }
 219
 220 static inline int8_t ff_u8_to_s8(uint8_t a)
 221 {
 222     union {
 223         uint8_t u8;
 224         int8_t  s8;
 225     } b;
 226     b.u8 = a;
 227     return b.s8;
 228 }
 229
 230 #endif /* AVCODEC_MATHOPS_H */