view softfloat.h @ 728:1fa3820b1a84 libavutil

ARM asm for AV_RN*() ARMv6 and later support unaligned loads and stores for single word/halfword but not double/multiple. GCC is ignorant of this and will always use bytewise accesses for unaligned data. Casting to an int32_t pointer is dangerous since a load/store double or multiple instruction might be used (this happens with some code in FFmpeg). Implementing the AV_[RW]* macros with inline asm using only supported instructions gives fast and safe unaligned accesses. ARM RVCT does the right thing with generic code. This gives an overall speedup of up to 10%.

/*
 * Copyright (c) 2006 Michael Niedermayer <[e-mail]>
 *
 * This file is part of FFmpeg.
 *
 * FFmpeg is free software; you can redistribute it and/or
 * modify it under the terms of the GNU Lesser General Public
 * License as published by the Free Software Foundation; either
 * version 2.1 of the License, or (at your option) any later version.
 *
 * FFmpeg is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
 * Lesser General Public License for more details.
 *
 * You should have received a copy of the GNU Lesser General Public
 * License along with FFmpeg; if not, write to the Free Software
 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
 */

#ifndef AVUTIL_SOFTFLOAT_H
#define AVUTIL_SOFTFLOAT_H

#include <stdint.h>
#include "common.h"

#define MIN_EXP -126
#define MAX_EXP  126
#define ONE_BITS 29

typedef struct SoftFloat{
    int32_t  exp;
    int32_t mant;
}SoftFloat;

static av_const SoftFloat av_normalize_sf(SoftFloat a){
    if(a.mant){
#if 1
        while((a.mant + 0x20000000U)<0x40000000U){
            a.mant += a.mant;
            a.exp  -= 1;
        }
#else
        int s=ONE_BITS + 1 - av_log2(a.mant ^ (a.mant<<1));
        a.exp   -= s;
        a.mant <<= s;
#endif
        if(a.exp < MIN_EXP){
            a.exp = MIN_EXP;
            a.mant= 0;
        }
    }else{
        a.exp= MIN_EXP;
    }
    return a;
}

static inline av_const SoftFloat av_normalize1_sf(SoftFloat a){
#if 1
    if(a.mant + 0x40000000 < 0){
        a.exp++;
        a.mant>>=1;
    }
    return a;
#elif 1
    int t= a.mant + 0x40000000 < 0;
    return (SoftFloat){a.exp+t, a.mant>>t};
#else
    int t= (a.mant + 0x40000000U)>>31;
    return (SoftFloat){a.exp+t, a.mant>>t};
#endif
}

/**
 * @return Will not be more denormalized than a+b. So if either input is
 *         normalized, then the output will not be worse then the other input.
 *         If both are normalized, then the output will be normalized.
 */
static inline av_const SoftFloat av_mul_sf(SoftFloat a, SoftFloat b){
    a.exp += b.exp;
    a.mant = (a.mant * (int64_t)b.mant) >> ONE_BITS;
    return av_normalize1_sf(a);
}

/**
 * b has to be normalized and not zero.
 * @return Will not be more denormalized than a.
 */
static av_const SoftFloat av_div_sf(SoftFloat a, SoftFloat b){
    a.exp -= b.exp+1;
    a.mant = ((int64_t)a.mant<<(ONE_BITS+1)) / b.mant;
    return av_normalize1_sf(a);
}

static inline av_const int av_cmp_sf(SoftFloat a, SoftFloat b){
    int t= a.exp - b.exp;
    if(t<0) return (a.mant >> (-t)) -  b.mant      ;
    else    return  a.mant          - (b.mant >> t);
}

static inline av_const SoftFloat av_add_sf(SoftFloat a, SoftFloat b){
    int t= a.exp - b.exp;
    if(t<0) return av_normalize1_sf((SoftFloat){b.exp, b.mant + (a.mant >> (-t))});
    else    return av_normalize1_sf((SoftFloat){a.exp, a.mant + (b.mant >>   t )});
}

static inline av_const SoftFloat av_sub_sf(SoftFloat a, SoftFloat b){
    return av_add_sf(a, (SoftFloat){b.exp, -b.mant});
}

//FIXME sqrt, log, exp, pow, sin, cos

static inline av_const SoftFloat av_int2sf(int v, int frac_bits){
    return av_normalize_sf((SoftFloat){ONE_BITS-frac_bits, v});
}

/**
 * Rounding is to -inf.
 */
static inline av_const int av_sf2int(SoftFloat v, int frac_bits){
    v.exp += frac_bits - ONE_BITS;
    if(v.exp >= 0) return v.mant <<  v.exp ;
    else           return v.mant >>(-v.exp);
}

#endif /* AVUTIL_SOFTFLOAT_H */