X-Git-Url: https://git.sesse.net/?a=blobdiff_plain;f=libavcodec%2Farm%2Fmathops.h;h=45ac67d43625515678c53d05fdee98f85688267e;hb=efd26bedec9a345a5960dbfcbaec888418f2d4e6;hp=e36316c76bed7ba1412819fcb448711bfe03b4c2;hpb=a2fc0f6a6ddf884ace3c96a0d4f09f0932e6db32;p=ffmpeg diff --git a/libavcodec/arm/mathops.h b/libavcodec/arm/mathops.h index e36316c76be..45ac67d4362 100644 --- a/libavcodec/arm/mathops.h +++ b/libavcodec/arm/mathops.h @@ -2,20 +2,20 @@ * simple math operations * Copyright (c) 2006 Michael Niedermayer et al * - * This file is part of FFmpeg. + * This file is part of Libav. * - * FFmpeg is free software; you can redistribute it and/or + * Libav 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, + * Libav 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 + * License along with Libav; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA */ @@ -23,57 +23,48 @@ #define AVCODEC_ARM_MATHOPS_H #include +#include "config.h" #include "libavutil/common.h" -# define MULL MULL -static inline av_const int MULL(int a, int b, unsigned shift) -{ - int lo, hi; - __asm__("smull %0, %1, %2, %3 \n\t" - "mov %0, %0, lsr %4 \n\t" - "add %1, %0, %1, lsl %5 \n\t" - : "=&r"(lo), "=&r"(hi) - : "r"(b), "r"(a), "i"(shift), "i"(32-shift)); - return hi; -} +#if HAVE_INLINE_ASM +#if HAVE_ARMV6_INLINE #define MULH MULH -#ifdef HAVE_ARMV6 static inline av_const int MULH(int a, int b) { int r; __asm__ ("smmul %0, %1, %2" : "=r"(r) : "r"(a), "r"(b)); return r; } -#else -static inline av_const int MULH(int a, int b) -{ - int lo, hi; - __asm__ ("smull %0, %1, %2, %3" : "=&r"(lo), "=&r"(hi) : "r"(b), "r"(a)); - return hi; -} -#endif -static inline av_const int64_t MUL64(int a, int b) +#define FASTDIV FASTDIV +static av_always_inline av_const int FASTDIV(int a, int b) { - union { uint64_t x; unsigned hl[2]; } x; - __asm__ ("smull %0, %1, %2, %3" - : "=r"(x.hl[0]), "=r"(x.hl[1]) : "r"(a), "r"(b)); - return x.x; + int r; + __asm__ ("cmp %2, #2 \n\t" + "ldr %0, [%3, %2, lsl #2] \n\t" + "ite le \n\t" + "lsrle %0, %1, #1 \n\t" + "smmulgt %0, %0, %1 \n\t" + : "=&r"(r) : "r"(a), "r"(b), "r"(ff_inverse) : "cc"); + return r; } -#define MUL64 MUL64 -static inline av_const int64_t MAC64(int64_t d, int a, int b) +#else /* HAVE_ARMV6_INLINE */ + +#define FASTDIV FASTDIV +static av_always_inline av_const int FASTDIV(int a, int b) { - union { uint64_t x; unsigned hl[2]; } x = { d }; - __asm__ ("smlal %0, %1, %2, %3" - : "+r"(x.hl[0]), "+r"(x.hl[1]) : "r"(a), "r"(b)); - return x.x; + int r, t; + __asm__ ("umull %1, %0, %2, %3" + : "=&r"(r), "=&r"(t) : "r"(a), "r"(ff_inverse[b])); + return r; } -#define MAC64(d, a, b) ((d) = MAC64(d, a, b)) +#endif + #define MLS64(d, a, b) MAC64(d, -(a), b) -#if defined(HAVE_ARMV5TE) +#if HAVE_ARMV5TE_INLINE /* signed 16x16 -> 32 multiply add accumulate */ # define MAC16(rt, ra, rb) \ @@ -81,7 +72,7 @@ static inline av_const int64_t MAC64(int64_t d, int a, int b) /* signed 16x16 -> 32 multiply */ # define MUL16 MUL16 -static inline av_const MUL16(int ra, int rb) +static inline av_const int MUL16(int ra, int rb) { int rt; __asm__ ("smulbb %0, %1, %2" : "=r"(rt) : "r"(ra), "r"(rb)); @@ -90,4 +81,28 @@ static inline av_const MUL16(int ra, int rb) #endif +#define mid_pred mid_pred +static inline av_const int mid_pred(int a, int b, int c) +{ + int m; + __asm__ ( + "mov %0, %2 \n\t" + "cmp %1, %2 \n\t" + "itt gt \n\t" + "movgt %0, %1 \n\t" + "movgt %1, %2 \n\t" + "cmp %1, %3 \n\t" + "it le \n\t" + "movle %1, %3 \n\t" + "cmp %0, %1 \n\t" + "it gt \n\t" + "movgt %0, %1 \n\t" + : "=&r"(m), "+r"(a) + : "r"(b), "r"(c) + : "cc"); + return m; +} + +#endif /* HAVE_INLINE_ASM */ + #endif /* AVCODEC_ARM_MATHOPS_H */