vp8/encoder/arm/neon/sad8_neon.asm

   1 ;
   2 ;  Copyright (c) 2010 The WebM project authors. All Rights Reserved.
   3 ;
   4 ;  Use of this source code is governed by a BSD-style license
   5 ;  that can be found in the LICENSE file in the root of the source
   6 ;  tree. An additional intellectual property rights grant can be found
   7 ;  in the file PATENTS.  All contributing project authors may
   8 ;  be found in the AUTHORS file in the root of the source tree.
   9 ;
  10
  11
  12     EXPORT  |vp8_sad8x8_neon|
  13     EXPORT  |vp8_sad8x16_neon|
  14     EXPORT  |vp8_sad4x4_neon|
  15
  16     ARM
  17     REQUIRE8
  18     PRESERVE8
  19
  20     AREA ||.text||, CODE, READONLY, ALIGN=2
  21 ; unsigned int vp8_sad8x8_c(
  22 ;    unsigned char *src_ptr,
  23 ;    int  src_stride,
  24 ;    unsigned char *ref_ptr,
  25 ;    int  ref_stride)
  26
  27 |vp8_sad8x8_neon| PROC
  28     vld1.8          {d0}, [r0], r1
  29     vld1.8          {d8}, [r2], r3
  30
  31     vld1.8          {d2}, [r0], r1
  32     vld1.8          {d10}, [r2], r3
  33
  34     vabdl.u8        q12, d0, d8
  35
  36     vld1.8          {d4}, [r0], r1
  37     vld1.8          {d12}, [r2], r3
  38
  39     vabal.u8        q12, d2, d10
  40
  41     vld1.8          {d6}, [r0], r1
  42     vld1.8          {d14}, [r2], r3
  43
  44     vabal.u8        q12, d4, d12
  45
  46     vld1.8          {d0}, [r0], r1
  47     vld1.8          {d8}, [r2], r3
  48
  49     vabal.u8        q12, d6, d14
  50
  51     vld1.8          {d2}, [r0], r1
  52     vld1.8          {d10}, [r2], r3
  53
  54     vabal.u8        q12, d0, d8
  55
  56     vld1.8          {d4}, [r0], r1
  57     vld1.8          {d12}, [r2], r3
  58
  59     vabal.u8        q12, d2, d10
  60
  61     vld1.8          {d6}, [r0], r1
  62     vld1.8          {d14}, [r2], r3
  63
  64     vabal.u8        q12, d4, d12
  65     vabal.u8        q12, d6, d14
  66
  67     vpaddl.u16      q1, q12
  68     vpaddl.u32      q0, q1
  69     vadd.u32        d0, d0, d1
  70
  71     vmov.32         r0, d0[0]
  72
  73     bx              lr
  74
  75     ENDP
  76
  77 ;============================
  78 ;unsigned int vp8_sad8x16_c(
  79 ;    unsigned char *src_ptr,
  80 ;    int  src_stride,
  81 ;    unsigned char *ref_ptr,
  82 ;    int  ref_stride)
  83
  84 |vp8_sad8x16_neon| PROC
  85     vld1.8          {d0}, [r0], r1
  86     vld1.8          {d8}, [r2], r3
  87
  88     vld1.8          {d2}, [r0], r1
  89     vld1.8          {d10}, [r2], r3
  90
  91     vabdl.u8        q12, d0, d8
  92
  93     vld1.8          {d4}, [r0], r1
  94     vld1.8          {d12}, [r2], r3
  95
  96     vabal.u8        q12, d2, d10
  97
  98     vld1.8          {d6}, [r0], r1
  99     vld1.8          {d14}, [r2], r3
 100
 101     vabal.u8        q12, d4, d12
 102
 103     vld1.8          {d0}, [r0], r1
 104     vld1.8          {d8}, [r2], r3
 105
 106     vabal.u8        q12, d6, d14
 107
 108     vld1.8          {d2}, [r0], r1
 109     vld1.8          {d10}, [r2], r3
 110
 111     vabal.u8        q12, d0, d8
 112
 113     vld1.8          {d4}, [r0], r1
 114     vld1.8          {d12}, [r2], r3
 115
 116     vabal.u8        q12, d2, d10
 117
 118     vld1.8          {d6}, [r0], r1
 119     vld1.8          {d14}, [r2], r3
 120
 121     vabal.u8        q12, d4, d12
 122
 123     vld1.8          {d0}, [r0], r1
 124     vld1.8          {d8}, [r2], r3
 125
 126     vabal.u8        q12, d6, d14
 127
 128     vld1.8          {d2}, [r0], r1
 129     vld1.8          {d10}, [r2], r3
 130
 131     vabal.u8        q12, d0, d8
 132
 133     vld1.8          {d4}, [r0], r1
 134     vld1.8          {d12}, [r2], r3
 135
 136     vabal.u8        q12, d2, d10
 137
 138     vld1.8          {d6}, [r0], r1
 139     vld1.8          {d14}, [r2], r3
 140
 141     vabal.u8        q12, d4, d12
 142
 143     vld1.8          {d0}, [r0], r1
 144     vld1.8          {d8}, [r2], r3
 145
 146     vabal.u8        q12, d6, d14
 147
 148     vld1.8          {d2}, [r0], r1
 149     vld1.8          {d10}, [r2], r3
 150
 151     vabal.u8        q12, d0, d8
 152
 153     vld1.8          {d4}, [r0], r1
 154     vld1.8          {d12}, [r2], r3
 155
 156     vabal.u8        q12, d2, d10
 157
 158     vld1.8          {d6}, [r0], r1
 159     vld1.8          {d14}, [r2], r3
 160
 161     vabal.u8        q12, d4, d12
 162     vabal.u8        q12, d6, d14
 163
 164     vpaddl.u16      q1, q12
 165     vpaddl.u32      q0, q1
 166     vadd.u32        d0, d0, d1
 167
 168     vmov.32         r0, d0[0]
 169
 170     bx              lr
 171
 172     ENDP
 173
 174 ;===========================
 175 ;unsigned int vp8_sad4x4_c(
 176 ;    unsigned char *src_ptr,
 177 ;    int  src_stride,
 178 ;    unsigned char *ref_ptr,
 179 ;    int  ref_stride)
 180
 181 |vp8_sad4x4_neon| PROC
 182     vld1.8          {d0}, [r0], r1
 183     vld1.8          {d8}, [r2], r3
 184
 185     vld1.8          {d2}, [r0], r1
 186     vld1.8          {d10}, [r2], r3
 187
 188     vabdl.u8        q12, d0, d8
 189
 190     vld1.8          {d4}, [r0], r1
 191     vld1.8          {d12}, [r2], r3
 192
 193     vabal.u8        q12, d2, d10
 194
 195     vld1.8          {d6}, [r0], r1
 196     vld1.8          {d14}, [r2], r3
 197
 198     vabal.u8        q12, d4, d12
 199     vabal.u8        q12, d6, d14
 200
 201     vpaddl.u16      d1, d24
 202     vpaddl.u32      d0, d1
 203     vmov.32         r0, d0[0]
 204
 205     bx              lr
 206
 207     ENDP
 208
 209     END