libavcodec/arm/mathops.h

   1 /*
   2  * simple math operations
   3  * Copyright (c) 2006 Michael Niedermayer <michaelni@gmx.at> et al
   4  *
   5  * This file is part of FFmpeg.
   6  *
   7  * FFmpeg is free software; you can redistribute it and/or
   8  * modify it under the terms of the GNU Lesser General Public
   9  * License as published by the Free Software Foundation; either
  10  * version 2.1 of the License, or (at your option) any later version.
  11  *
  12  * FFmpeg is distributed in the hope that it will be useful,
  13  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  14  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  15  * Lesser General Public License for more details.
  16  *
  17  * You should have received a copy of the GNU Lesser General Public
  18  * License along with FFmpeg; if not, write to the Free Software
  19  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
  20  */
  21
  22 #ifndef AVCODEC_ARM_MATHOPS_H
  23 #define AVCODEC_ARM_MATHOPS_H
  24
  25 #include <stdint.h>
  26 #include "config.h"
  27 #include "libavutil/common.h"
  28
  29 #if HAVE_INLINE_ASM
  30
  31 #   define MULL MULL
  32 static inline av_const int MULL(int a, int b, unsigned shift)
  33 {
  34     int lo, hi;
  35     __asm__("smull %0, %1, %2, %3     \n\t"
  36             "mov   %0, %0,     lsr %4 \n\t"
  37             "add   %1, %0, %1, lsl %5 \n\t"
  38             : "=&r"(lo), "=&r"(hi)
  39             : "r"(b), "r"(a), "ir"(shift), "ir"(32-shift));
  40     return hi;
  41 }
  42
  43 #define MULH MULH
  44 #if HAVE_ARMV6
  45 static inline av_const int MULH(int a, int b)
  46 {
  47     int r;
  48     __asm__ ("smmul %0, %1, %2" : "=r"(r) : "r"(a), "r"(b));
  49     return r;
  50 }
  51 #else
  52 static inline av_const int MULH(int a, int b)
  53 {
  54     int lo, hi;
  55     __asm__ ("smull %0, %1, %2, %3" : "=&r"(lo), "=&r"(hi) : "r"(b), "r"(a));
  56     return hi;
  57 }
  58 #endif
  59
  60 static inline av_const int64_t MUL64(int a, int b)
  61 {
  62     union { uint64_t x; unsigned hl[2]; } x;
  63     __asm__ ("smull %0, %1, %2, %3"
  64              : "=r"(x.hl[0]), "=r"(x.hl[1]) : "r"(a), "r"(b));
  65     return x.x;
  66 }
  67 #define MUL64 MUL64
  68
  69 static inline av_const int64_t MAC64(int64_t d, int a, int b)
  70 {
  71     union { uint64_t x; unsigned hl[2]; } x = { d };
  72     __asm__ ("smlal %0, %1, %2, %3"
  73              : "+r"(x.hl[0]), "+r"(x.hl[1]) : "r"(a), "r"(b));
  74     return x.x;
  75 }
  76 #define MAC64(d, a, b) ((d) = MAC64(d, a, b))
  77 #define MLS64(d, a, b) MAC64(d, -(a), b)
  78
  79 #if HAVE_ARMV5TE
  80
  81 /* signed 16x16 -> 32 multiply add accumulate */
  82 #   define MAC16(rt, ra, rb)                                            \
  83     __asm__ ("smlabb %0, %1, %2, %0" : "+r"(rt) : "r"(ra), "r"(rb));
  84
  85 /* signed 16x16 -> 32 multiply */
  86 #   define MUL16 MUL16
  87 static inline av_const int MUL16(int ra, int rb)
  88 {
  89     int rt;
  90     __asm__ ("smulbb %0, %1, %2" : "=r"(rt) : "r"(ra), "r"(rb));
  91     return rt;
  92 }
  93
  94 #endif
  95
  96 #define mid_pred mid_pred
  97 static inline av_const int mid_pred(int a, int b, int c)
  98 {
  99     int m;
 100     __asm__ volatile (
 101         "mov   %0, %2  \n\t"
 102         "cmp   %1, %2  \n\t"
 103         "movgt %0, %1  \n\t"
 104         "movgt %1, %2  \n\t"
 105         "cmp   %1, %3  \n\t"
 106         "movle %1, %3  \n\t"
 107         "cmp   %0, %1  \n\t"
 108         "movgt %0, %1  \n\t"
 109         : "=&r"(m), "+r"(a)
 110         : "r"(b), "r"(c));
 111     return m;
 112 }
 113
 114 #endif /* HAVE_INLINE_ASM */
 115
 116 #endif /* AVCODEC_ARM_MATHOPS_H */