diff libavcodec/ppc/mathops.h @ 2:897f711a7157

rearrange to work with autoconf
author Nina Engelhardt <nengel@mailbox.tu-berlin.de>
date Tue, 25 Sep 2012 15:55:33 +0200
parents
children
line diff
     1.1 --- /dev/null	Thu Jan 01 00:00:00 1970 +0000
     1.2 +++ b/libavcodec/ppc/mathops.h	Tue Sep 25 15:55:33 2012 +0200
     1.3 @@ -0,0 +1,79 @@
     1.4 +/*
     1.5 + * simple math operations
     1.6 + * Copyright (c) 2001, 2002 Fabrice Bellard
     1.7 + * Copyright (c) 2006 Michael Niedermayer <michaelni@gmx.at> et al
     1.8 + *
     1.9 + * This file is part of FFmpeg.
    1.10 + *
    1.11 + * FFmpeg is free software; you can redistribute it and/or
    1.12 + * modify it under the terms of the GNU Lesser General Public
    1.13 + * License as published by the Free Software Foundation; either
    1.14 + * version 2.1 of the License, or (at your option) any later version.
    1.15 + *
    1.16 + * FFmpeg is distributed in the hope that it will be useful,
    1.17 + * but WITHOUT ANY WARRANTY; without even the implied warranty of
    1.18 + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
    1.19 + * Lesser General Public License for more details.
    1.20 + *
    1.21 + * You should have received a copy of the GNU Lesser General Public
    1.22 + * License along with FFmpeg; if not, write to the Free Software
    1.23 + * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
    1.24 + */
    1.25 +
    1.26 +#ifndef AVCODEC_PPC_MATHOPS_H
    1.27 +#define AVCODEC_PPC_MATHOPS_H
    1.28 +
    1.29 +#include <stdint.h>
    1.30 +#include "config.h"
    1.31 +#include "libavutil/common.h"
    1.32 +
    1.33 +#if HAVE_PPC4XX
    1.34 +/* signed 16x16 -> 32 multiply add accumulate */
    1.35 +#define MAC16(rt, ra, rb) \
    1.36 +    __asm__ ("maclhw %0, %2, %3" : "=r" (rt) : "0" (rt), "r" (ra), "r" (rb));
    1.37 +
    1.38 +/* signed 16x16 -> 32 multiply */
    1.39 +#define MUL16(ra, rb) \
    1.40 +    ({ int __rt; \
    1.41 +    __asm__ ("mullhw %0, %1, %2" : "=r" (__rt) : "r" (ra), "r" (rb)); \
    1.42 +    __rt; })
    1.43 +#endif
    1.44 +
    1.45 +#define MULH MULH
    1.46 +static inline av_const int MULH(int a, int b){
    1.47 +    int r;
    1.48 +    __asm__ ("mulhw %0, %1, %2" : "=r"(r) : "r"(a), "r"(b));
    1.49 +    return r;
    1.50 +}
    1.51 +
    1.52 +#if !ARCH_PPC64
    1.53 +static inline av_const int64_t MAC64(int64_t d, int a, int b)
    1.54 +{
    1.55 +    union { uint64_t x; unsigned hl[2]; } x = { d };
    1.56 +    int h, l;
    1.57 +    __asm__ ("mullw %3, %4, %5   \n\t"
    1.58 +             "mulhw %2, %4, %5   \n\t"
    1.59 +             "addc  %1, %1, %3   \n\t"
    1.60 +             "adde  %0, %0, %2   \n\t"
    1.61 +             : "+r"(x.hl[0]), "+r"(x.hl[1]), "=&r"(h), "=&r"(l)
    1.62 +             : "r"(a), "r"(b));
    1.63 +    return x.x;
    1.64 +}
    1.65 +#define MAC64(d, a, b) ((d) = MAC64(d, a, b))
    1.66 +
    1.67 +static inline av_const int64_t MLS64(int64_t d, int a, int b)
    1.68 +{
    1.69 +    union { uint64_t x; unsigned hl[2]; } x = { d };
    1.70 +    int h, l;
    1.71 +    __asm__ ("mullw %3, %4, %5   \n\t"
    1.72 +             "mulhw %2, %4, %5   \n\t"
    1.73 +             "subfc %1, %3, %1   \n\t"
    1.74 +             "subfe %0, %2, %0   \n\t"
    1.75 +             : "+r"(x.hl[0]), "+r"(x.hl[1]), "=&r"(h), "=&r"(l)
    1.76 +             : "r"(a), "r"(b));
    1.77 +    return x.x;
    1.78 +}
    1.79 +#define MLS64(d, a, b) ((d) = MLS64(d, a, b))
    1.80 +#endif
    1.81 +
    1.82 +#endif /* AVCODEC_PPC_MATHOPS_H */