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"
30 #define MAX_NEG_CROP 1024
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];
39 # include "arm/mathops.h"
41 # include "avr32/mathops.h"
43 # include "mips/mathops.h"
45 # include "ppc/mathops.h"
47 # include "x86/mathops.h"
50 /* generic implementation */
53 # define MUL64(a,b) ((int64_t)(a) * (int64_t)(b))
57 # define MULL(a,b,s) (MUL64(a, b) >> (s))
61 static av_always_inline int MULH(int a, int b){
62 return MUL64(a, b) >> 32;
67 static av_always_inline unsigned UMULH(unsigned a, unsigned b){
68 return ((uint64_t)(a) * (uint64_t)(b))>>32;
73 # define MAC64(d, a, b) ((d) += MUL64(a, b))
77 # define MLS64(d, a, b) ((d) -= MUL64(a, b))
80 /* signed 16x16 -> 32 multiply add accumulate */
82 # define MAC16(rt, ra, rb) rt += (ra) * (rb)
85 /* signed 16x16 -> 32 multiply */
87 # define MUL16(ra, rb) ((ra) * (rb))
91 # define MLS16(rt, ra, rb) ((rt) -= (ra) * (rb))
96 #define mid_pred mid_pred
97 static inline av_const int mid_pred(int a, int b, int c)
100 int t= (a-b)&((a-b)>>31);
103 b-= (b-c)&((b-c)>>31);
104 b+= (a-b)&((a-b)>>31);
125 #define median4 median4
126 static inline av_const int median4(int a, int b, int c, int d)
129 if (c < d) return (FFMIN(b, d) + FFMAX(a, c)) / 2;
130 else return (FFMIN(b, c) + FFMAX(a, d)) / 2;
132 if (c < d) return (FFMIN(a, d) + FFMAX(b, c)) / 2;
133 else return (FFMIN(a, c) + FFMAX(b, d)) / 2;
139 static inline av_const int sign_extend(int val, unsigned bits)
141 unsigned shift = 8 * sizeof(int) - bits;
142 union { unsigned u; int s; } v = { (unsigned) val << shift };
148 static inline av_const unsigned zero_extend(unsigned val, unsigned bits)
150 return (val << ((8 * sizeof(int)) - bits)) >> ((8 * sizeof(int)) - bits);
155 #define COPY3_IF_LT(x, y, a, b, c, d)\
164 #define MASK_ABS(mask, level) do { \
165 mask = level >> 31; \
166 level = (level ^ mask) - mask; \
171 # define NEG_SSR32(a,s) ((( int32_t)(a))>>(32-(s)))
175 # define NEG_USR32(a,s) (((uint32_t)(a))>>(32-(s)))
180 # define PACK_2U8(a,b) (((a) << 8) | (b))
183 # define PACK_4U8(a,b,c,d) (((a) << 24) | ((b) << 16) | ((c) << 8) | (d))
186 # define PACK_2U16(a,b) (((a) << 16) | (b))
190 # define PACK_2U8(a,b) (((b) << 8) | (a))
193 # define PACK_4U8(a,b,c,d) (((d) << 24) | ((c) << 16) | ((b) << 8) | (a))
196 # define PACK_2U16(a,b) (((b) << 16) | (a))
201 # define PACK_2S8(a,b) PACK_2U8((a)&255, (b)&255)
204 # define PACK_4S8(a,b,c,d) PACK_4U8((a)&255, (b)&255, (c)&255, (d)&255)
207 # define PACK_2S16(a,b) PACK_2U16((a)&0xffff, (b)&0xffff)
211 # define FASTDIV(a,b) ((uint32_t)((((uint64_t)a) * ff_inverse[b]) >> 32))
214 static inline av_const unsigned int ff_sqrt(unsigned int a)
218 if (a < 255) return (ff_sqrt_tab[a + 1] - 1) >> 4;
219 else if (a < (1 << 12)) b = ff_sqrt_tab[a >> 4] >> 2;
221 else if (a < (1 << 14)) b = ff_sqrt_tab[a >> 6] >> 1;
222 else if (a < (1 << 16)) b = ff_sqrt_tab[a >> 8] ;
225 int s = av_log2_16bit(a >> 16) >> 1;
226 unsigned int c = a >> (s + 2);
227 b = ff_sqrt_tab[c >> (s + 8)];
228 b = FASTDIV(c,b) + (b << s);
231 return b - (a < b * b);
234 static inline int8_t ff_u8_to_s8(uint8_t a)
244 #endif /* AVCODEC_MATHOPS_H */