2 * simple math operations
3 * Copyright (c) 2001, 2002 Fabrice Bellard
4 * Copyright (c) 2006 Michael Niedermayer <michaelni@gmx.at> et al
6 * This file is part of FFmpeg.
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.
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.
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
22 #ifndef AVCODEC_MATHOPS_H
23 #define AVCODEC_MATHOPS_H
27 #include "libavutil/common.h"
28 #include "libavutil/reverse.h"
31 #define MAX_NEG_CROP 1024
33 extern const uint32_t ff_inverse[257];
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 extern const uint8_t ff_zigzag_scan[16+1];
40 # include "arm/mathops.h"
42 # include "avr32/mathops.h"
44 # include "mips/mathops.h"
46 # include "ppc/mathops.h"
48 # include "x86/mathops.h"
51 /* generic implementation */
54 # define MUL64(a,b) ((int64_t)(a) * (int64_t)(b))
58 # define MULL(a,b,s) (MUL64(a, b) >> (s))
62 static av_always_inline int MULH(int a, int b){
63 return MUL64(a, b) >> 32;
68 static av_always_inline unsigned UMULH(unsigned a, unsigned b){
69 return ((uint64_t)(a) * (uint64_t)(b))>>32;
74 # define MAC64(d, a, b) ((d) += MUL64(a, b))
78 # define MLS64(d, a, b) ((d) -= MUL64(a, b))
81 /* signed 16x16 -> 32 multiply add accumulate */
83 # define MAC16(rt, ra, rb) rt += (ra) * (rb)
86 /* signed 16x16 -> 32 multiply */
88 # define MUL16(ra, rb) ((ra) * (rb))
92 # define MLS16(rt, ra, rb) ((rt) -= (ra) * (rb))
97 #define mid_pred mid_pred
98 static inline av_const int mid_pred(int a, int b, int c)
116 #define median4 median4
117 static inline av_const int median4(int a, int b, int c, int d)
120 if (c < d) return (FFMIN(b, d) + FFMAX(a, c)) / 2;
121 else return (FFMIN(b, c) + FFMAX(a, d)) / 2;
123 if (c < d) return (FFMIN(a, d) + FFMAX(b, c)) / 2;
124 else return (FFMIN(a, c) + FFMAX(b, d)) / 2;
130 static inline av_const int sign_extend(int val, unsigned bits)
132 unsigned shift = 8 * sizeof(int) - bits;
133 union { unsigned u; int s; } v = { (unsigned) val << shift };
139 static inline av_const unsigned zero_extend(unsigned val, unsigned bits)
141 return (val << ((8 * sizeof(int)) - bits)) >> ((8 * sizeof(int)) - bits);
146 #define COPY3_IF_LT(x, y, a, b, c, d)\
155 #define MASK_ABS(mask, level) do { \
156 mask = level >> 31; \
157 level = (level ^ mask) - mask; \
162 # define NEG_SSR32(a,s) ((( int32_t)(a))>>(32-(s)))
166 # define NEG_USR32(a,s) (((uint32_t)(a))>>(32-(s)))
171 # define PACK_2U8(a,b) (((a) << 8) | (b))
174 # define PACK_4U8(a,b,c,d) (((a) << 24) | ((b) << 16) | ((c) << 8) | (d))
177 # define PACK_2U16(a,b) (((a) << 16) | (b))
181 # define PACK_2U8(a,b) (((b) << 8) | (a))
184 # define PACK_4U8(a,b,c,d) (((d) << 24) | ((c) << 16) | ((b) << 8) | (a))
187 # define PACK_2U16(a,b) (((b) << 16) | (a))
192 # define PACK_2S8(a,b) PACK_2U8((a)&255, (b)&255)
195 # define PACK_4S8(a,b,c,d) PACK_4U8((a)&255, (b)&255, (c)&255, (d)&255)
198 # define PACK_2S16(a,b) PACK_2U16((a)&0xffff, (b)&0xffff)
202 # define FASTDIV(a,b) ((uint32_t)((((uint64_t)a) * ff_inverse[b]) >> 32))
206 #define ff_sqrt ff_sqrt
207 static inline av_const unsigned int ff_sqrt(unsigned int a)
211 if (a < 255) return (ff_sqrt_tab[a + 1] - 1) >> 4;
212 else if (a < (1 << 12)) b = ff_sqrt_tab[a >> 4] >> 2;
214 else if (a < (1 << 14)) b = ff_sqrt_tab[a >> 6] >> 1;
215 else if (a < (1 << 16)) b = ff_sqrt_tab[a >> 8] ;
218 int s = av_log2_16bit(a >> 16) >> 1;
219 unsigned int c = a >> (s + 2);
220 b = ff_sqrt_tab[c >> (s + 8)];
221 b = FASTDIV(c,b) + (b << s);
224 return b - (a < b * b);
228 static inline av_const float ff_sqrf(float a)
233 static inline int8_t ff_u8_to_s8(uint8_t a)
243 static av_always_inline uint32_t bitswap_32(uint32_t x)
245 return (uint32_t)ff_reverse[ x & 0xFF] << 24 |
246 (uint32_t)ff_reverse[(x >> 8) & 0xFF] << 16 |
247 (uint32_t)ff_reverse[(x >> 16) & 0xFF] << 8 |
248 (uint32_t)ff_reverse[ x >> 24];
251 #endif /* AVCODEC_MATHOPS_H */