ffmpeg/libavcodec/x86/pixblockdsp.asm

   1 ;*****************************************************************************
   2 ;* SIMD-optimized pixel operations
   3 ;*****************************************************************************
   4 ;* Copyright (c) 2000, 2001 Fabrice Bellard
   5 ;* Copyright (c) 2002-2004 Michael Niedermayer <michaelni@gmx.at>
   6 ;*
   7 ;* This file is part of FFmpeg.
   8 ;*
   9 ;* FFmpeg is free software; you can redistribute it and/or
  10 ;* modify it under the terms of the GNU Lesser General Public
  11 ;* License as published by the Free Software Foundation; either
  12 ;* version 2.1 of the License, or (at your option) any later version.
  13 ;*
  14 ;* FFmpeg is distributed in the hope that it will be useful,
  15 ;* but WITHOUT ANY WARRANTY; without even the implied warranty of
  16 ;* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  17 ;* Lesser General Public License for more details.
  18 ;*
  19 ;* You should have received a copy of the GNU Lesser General Public
  20 ;* License along with FFmpeg; if not, write to the Free Software
  21 ;* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
  22 ;*****************************************************************************
  23
  24 %include "libavutil/x86/x86util.asm"
  25
  26 SECTION .text
  27
  28 INIT_MMX mmx
  29 ; void ff_get_pixels_mmx(int16_t *block, const uint8_t *pixels, ptrdiff_t line_size)
  30 cglobal get_pixels, 3,4
  31     add          r0, 128
  32     mov          r3, -128
  33     pxor         m7, m7
  34 .loop:
  35     mova         m0, [r1]
  36     mova         m2, [r1+r2]
  37     mova         m1, m0
  38     mova         m3, m2
  39     punpcklbw    m0, m7
  40     punpckhbw    m1, m7
  41     punpcklbw    m2, m7
  42     punpckhbw    m3, m7
  43     mova [r0+r3+ 0], m0
  44     mova [r0+r3+ 8], m1
  45     mova [r0+r3+16], m2
  46     mova [r0+r3+24], m3
  47     lea          r1, [r1+r2*2]
  48     add          r3, 32
  49     js .loop
  50     REP_RET
  51
  52 INIT_XMM sse2
  53 cglobal get_pixels, 3, 4, 5
  54     lea          r3, [r2*3]
  55     pxor         m4, m4
  56     movh         m0, [r1]
  57     movh         m1, [r1+r2]
  58     movh         m2, [r1+r2*2]
  59     movh         m3, [r1+r3]
  60     lea          r1, [r1+r2*4]
  61     punpcklbw    m0, m4
  62     punpcklbw    m1, m4
  63     punpcklbw    m2, m4
  64     punpcklbw    m3, m4
  65     mova       [r0], m0
  66     mova  [r0+0x10], m1
  67     mova  [r0+0x20], m2
  68     mova  [r0+0x30], m3
  69     movh         m0, [r1]
  70     movh         m1, [r1+r2*1]
  71     movh         m2, [r1+r2*2]
  72     movh         m3, [r1+r3]
  73     punpcklbw    m0, m4
  74     punpcklbw    m1, m4
  75     punpcklbw    m2, m4
  76     punpcklbw    m3, m4
  77     mova  [r0+0x40], m0
  78     mova  [r0+0x50], m1
  79     mova  [r0+0x60], m2
  80     mova  [r0+0x70], m3
  81     RET
  82
  83 INIT_MMX mmx
  84 ; void ff_diff_pixels_mmx(int16_t *block, const uint8_t *s1, const uint8_t *s2,
  85 ;                         int stride);
  86 cglobal diff_pixels, 4,5
  87     movsxdifnidn r3, r3d
  88     pxor         m7, m7
  89     add          r0,  128
  90     mov          r4, -128
  91 .loop:
  92     mova         m0, [r1]
  93     mova         m2, [r2]
  94     mova         m1, m0
  95     mova         m3, m2
  96     punpcklbw    m0, m7
  97     punpckhbw    m1, m7
  98     punpcklbw    m2, m7
  99     punpckhbw    m3, m7
 100     psubw        m0, m2
 101     psubw        m1, m3
 102     mova  [r0+r4+0], m0
 103     mova  [r0+r4+8], m1
 104     add          r1, r3
 105     add          r2, r3
 106     add          r4, 16
 107     jne .loop
 108     REP_RET
 109
 110 INIT_XMM sse2
 111 cglobal diff_pixels, 4, 5, 5
 112     movsxdifnidn r3, r3d
 113     pxor         m4, m4
 114     add          r0,  128
 115     mov          r4, -128
 116 .loop:
 117     movh         m0, [r1]
 118     movh         m2, [r2]
 119     movh         m1, [r1+r3]
 120     movh         m3, [r2+r3]
 121     punpcklbw    m0, m4
 122     punpcklbw    m1, m4
 123     punpcklbw    m2, m4
 124     punpcklbw    m3, m4
 125     psubw        m0, m2
 126     psubw        m1, m3
 127     mova [r0+r4+0 ], m0
 128     mova [r0+r4+16], m1
 129     lea          r1, [r1+r3*2]
 130     lea          r2, [r2+r3*2]
 131     add          r4, 32
 132     jne .loop
 133     RET