| Line | Branch | Exec | Source |
|---|---|---|---|
| 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/attributes_internal.h" | ||
| 28 | #include "libavutil/common.h" | ||
| 29 | #include "config.h" | ||
| 30 | |||
| 31 | #define MAX_NEG_CROP 1024 | ||
| 32 | |||
| 33 | extern const uint32_t ff_inverse[257]; | ||
| 34 | extern const uint8_t ff_log2_run[41]; | ||
| 35 | extern const uint8_t ff_sqrt_tab[256]; | ||
| 36 | extern const uint8_t attribute_visibility_hidden ff_crop_tab[256 + 2 * MAX_NEG_CROP]; | ||
| 37 | extern const uint8_t ff_zigzag_direct[64]; | ||
| 38 | extern const uint8_t ff_zigzag_scan[16+1]; | ||
| 39 | |||
| 40 | #if ARCH_ARM | ||
| 41 | # include "arm/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 | 47898309 | static av_always_inline int MULH(int a, int b){ | |
| 62 | 47898309 | return MUL64(a, b) >> 32; | |
| 63 | } | ||
| 64 | #endif | ||
| 65 | |||
| 66 | #ifndef UMULH | ||
| 67 | 499 | static av_always_inline unsigned UMULH(unsigned a, unsigned b){ | |
| 68 | 499 | 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(a>b){ | ||
| 100 | if(c>b){ | ||
| 101 | if(c>a) b=a; | ||
| 102 | else b=c; | ||
| 103 | } | ||
| 104 | }else{ | ||
| 105 | if(b>c){ | ||
| 106 | if(c>a) b=c; | ||
| 107 | else b=a; | ||
| 108 | } | ||
| 109 | } | ||
| 110 | return b; | ||
| 111 | } | ||
| 112 | #endif | ||
| 113 | |||
| 114 | #ifndef median4 | ||
| 115 | #define median4 median4 | ||
| 116 | 74348 | static inline av_const int median4(int a, int b, int c, int d) | |
| 117 | { | ||
| 118 |
2/2✓ Branch 0 taken 10403 times.
✓ Branch 1 taken 63945 times.
|
74348 | if (a < b) { |
| 119 |
2/2✓ Branch 0 taken 2971 times.
✓ Branch 1 taken 7432 times.
|
10403 | if (c < d) return (FFMIN(b, d) + FFMAX(a, c)) / 2; |
| 120 | 7432 | else return (FFMIN(b, c) + FFMAX(a, d)) / 2; | |
| 121 | } else { | ||
| 122 |
2/2✓ Branch 0 taken 5325 times.
✓ Branch 1 taken 58620 times.
|
63945 | if (c < d) return (FFMIN(a, d) + FFMAX(b, c)) / 2; |
| 123 | 58620 | else return (FFMIN(a, c) + FFMAX(b, d)) / 2; | |
| 124 | } | ||
| 125 | } | ||
| 126 | #endif | ||
| 127 | |||
| 128 | #define FF_SIGNBIT(x) ((x) >> CHAR_BIT * sizeof(x) - 1) | ||
| 129 | |||
| 130 | #ifndef sign_extend | ||
| 131 | 417773276 | static inline av_const int sign_extend(int val, unsigned bits) | |
| 132 | { | ||
| 133 | 417773276 | unsigned shift = 8 * sizeof(int) - bits; | |
| 134 | 417773276 | union { unsigned u; int s; } v = { (unsigned) val << shift }; | |
| 135 | 417773276 | return v.s >> shift; | |
| 136 | } | ||
| 137 | #endif | ||
| 138 | |||
| 139 | #ifndef sign_extend64 | ||
| 140 | 5648 | static inline av_const int64_t sign_extend64(int64_t val, unsigned bits) | |
| 141 | { | ||
| 142 | 5648 | unsigned shift = 8 * sizeof(int64_t) - bits; | |
| 143 | 5648 | union { uint64_t u; int64_t s; } v = { (uint64_t) val << shift }; | |
| 144 | 5648 | return v.s >> shift; | |
| 145 | } | ||
| 146 | #endif | ||
| 147 | |||
| 148 | #ifndef zero_extend | ||
| 149 | 80315433 | static inline av_const unsigned zero_extend(unsigned val, unsigned bits) | |
| 150 | { | ||
| 151 | 80315433 | return (val << ((8 * sizeof(int)) - bits)) >> ((8 * sizeof(int)) - bits); | |
| 152 | } | ||
| 153 | #endif | ||
| 154 | |||
| 155 | #ifndef COPY3_IF_LT | ||
| 156 | #define COPY3_IF_LT(x, y, a, b, c, d)\ | ||
| 157 | if ((y) < (x)) {\ | ||
| 158 | (x) = (y);\ | ||
| 159 | (a) = (b);\ | ||
| 160 | (c) = (d);\ | ||
| 161 | } | ||
| 162 | #endif | ||
| 163 | |||
| 164 | #ifndef MASK_ABS | ||
| 165 | #define MASK_ABS(mask, level) do { \ | ||
| 166 | mask = level >> 31; \ | ||
| 167 | level = (level ^ mask) - mask; \ | ||
| 168 | } while (0) | ||
| 169 | #endif | ||
| 170 | |||
| 171 | #ifndef NEG_SSR32 | ||
| 172 | # define NEG_SSR32(a,s) ((( int32_t)(a))>>(32-(s))) | ||
| 173 | #endif | ||
| 174 | |||
| 175 | #ifndef NEG_USR32 | ||
| 176 | # define NEG_USR32(a,s) (((uint32_t)(a))>>(32-(s))) | ||
| 177 | #endif | ||
| 178 | |||
| 179 | #if HAVE_BIGENDIAN | ||
| 180 | # ifndef PACK_2U8 | ||
| 181 | # define PACK_2U8(a,b) (((a) << 8) | (b)) | ||
| 182 | # endif | ||
| 183 | # ifndef PACK_4U8 | ||
| 184 | # define PACK_4U8(a,b,c,d) (((a) << 24) | ((b) << 16) | ((c) << 8) | (d)) | ||
| 185 | # endif | ||
| 186 | # ifndef PACK_2U16 | ||
| 187 | # define PACK_2U16(a,b) (((a) << 16) | (b)) | ||
| 188 | # endif | ||
| 189 | #else | ||
| 190 | # ifndef PACK_2U8 | ||
| 191 | # define PACK_2U8(a,b) (((b) << 8) | (a)) | ||
| 192 | # endif | ||
| 193 | # ifndef PACK_4U2 | ||
| 194 | # define PACK_4U8(a,b,c,d) (((d) << 24) | ((c) << 16) | ((b) << 8) | (a)) | ||
| 195 | # endif | ||
| 196 | # ifndef PACK_2U16 | ||
| 197 | # define PACK_2U16(a,b) (((b) << 16) | (a)) | ||
| 198 | # endif | ||
| 199 | #endif | ||
| 200 | |||
| 201 | #ifndef PACK_2S8 | ||
| 202 | # define PACK_2S8(a,b) PACK_2U8((a)&255, (b)&255) | ||
| 203 | #endif | ||
| 204 | #ifndef PACK_4S8 | ||
| 205 | # define PACK_4S8(a,b,c,d) PACK_4U8((a)&255, (b)&255, (c)&255, (d)&255) | ||
| 206 | #endif | ||
| 207 | #ifndef PACK_2S16 | ||
| 208 | # define PACK_2S16(a,b) PACK_2U16((a)&0xffff, (b)&0xffff) | ||
| 209 | #endif | ||
| 210 | |||
| 211 | #ifndef FASTDIV | ||
| 212 | # define FASTDIV(a,b) ((uint32_t)((((uint64_t)a) * ff_inverse[b]) >> 32)) | ||
| 213 | #endif /* FASTDIV */ | ||
| 214 | |||
| 215 | #ifndef ff_sqrt | ||
| 216 | #define ff_sqrt ff_sqrt | ||
| 217 | 7048311 | static inline av_const unsigned int ff_sqrt(unsigned int a) | |
| 218 | { | ||
| 219 | unsigned int b; | ||
| 220 | |||
| 221 |
2/2✓ Branch 0 taken 596599 times.
✓ Branch 1 taken 6451712 times.
|
7048311 | if (a < 255) return (ff_sqrt_tab[a + 1] - 1) >> 4; |
| 222 |
2/2✓ Branch 0 taken 847828 times.
✓ Branch 1 taken 5603884 times.
|
6451712 | else if (a < (1 << 12)) b = ff_sqrt_tab[a >> 4] >> 2; |
| 223 | #if !CONFIG_SMALL | ||
| 224 |
2/2✓ Branch 0 taken 1054425 times.
✓ Branch 1 taken 4549459 times.
|
5603884 | else if (a < (1 << 14)) b = ff_sqrt_tab[a >> 6] >> 1; |
| 225 |
2/2✓ Branch 0 taken 1046498 times.
✓ Branch 1 taken 3502961 times.
|
4549459 | else if (a < (1 << 16)) b = ff_sqrt_tab[a >> 8] ; |
| 226 | #endif | ||
| 227 | else { | ||
| 228 | 3502961 | int s = av_log2_16bit(a >> 16) >> 1; | |
| 229 | 3502961 | unsigned int c = a >> (s + 2); | |
| 230 | 3502961 | b = ff_sqrt_tab[c >> (s + 8)]; | |
| 231 | 3502961 | b = FASTDIV(c,b) + (b << s); | |
| 232 | } | ||
| 233 | |||
| 234 | 6451712 | return b - (a < b * b); | |
| 235 | } | ||
| 236 | #endif | ||
| 237 | |||
| 238 | 30243 | static inline av_const float ff_sqrf(float a) | |
| 239 | { | ||
| 240 | 30243 | return a*a; | |
| 241 | } | ||
| 242 | |||
| 243 | 393216 | static inline int8_t ff_u8_to_s8(uint8_t a) | |
| 244 | { | ||
| 245 | union { | ||
| 246 | uint8_t u8; | ||
| 247 | int8_t s8; | ||
| 248 | } b; | ||
| 249 | 393216 | b.u8 = a; | |
| 250 | 393216 | return b.s8; | |
| 251 | } | ||
| 252 | |||
| 253 | #endif /* AVCODEC_MATHOPS_H */ | ||
| 254 |