libavcodec/arm/mpegvideo_neon.S

   1 /*
   2  * Copyright (c) 2010 Mans Rullgard
   3  *
   4  * This file is part of FFmpeg.
   5  *
   6  * FFmpeg is free software; you can redistribute it and/or
   7  * modify it under the terms of the GNU Lesser General Public
   8  * License as published by the Free Software Foundation; either
   9  * version 2.1 of the License, or (at your option) any later version.
  10  *
  11  * FFmpeg is distributed in the hope that it will be useful,
  12  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  13  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  14  * Lesser General Public License for more details.
  15  *
  16  * You should have received a copy of the GNU Lesser General Public
  17  * License along with FFmpeg; if not, write to the Free Software
  18  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
  19  */
  20
  21 #include "asm.S"
  22 #include "asm-offsets.h"
  23
  24 function ff_dct_unquantize_h263_inter_neon, export=1
  25         add             r0,  r0,  #0x2200
  26         add             r12, r0,  #BLOCK_LAST_INDEX-0x2200
  27         ldr             r12, [r12, r2, lsl #2]
  28         add             r0,  r0,  #INTER_SCANTAB_RASTER_END-0x2200
  29         ldrb            r12, [r0, r12]
  30         sub             r2,  r3,  #1
  31         lsl             r0,  r3,  #1
  32         orr             r2,  r2,  #1
  33         add             r3,  r12, #1
  34 endfunc
  35
  36 function ff_dct_unquantize_h263_neon, export=1
  37         vdup.16         q15, r0                 @ qmul
  38         vdup.16         q14, r2                 @ qadd
  39         vneg.s16        q13, q14
  40         cmp             r3,  #4
  41         mov             r0,  r1
  42         ble             2f
  43 1:
  44         vld1.16         {q0},     [r0,:128]!
  45         vclt.s16        q3,  q0,  #0
  46         vld1.16         {q8},     [r0,:128]!
  47         vceq.s16        q1,  q0,  #0
  48         vmul.s16        q2,  q0,  q15
  49         vclt.s16        q11, q8,  #0
  50         vmul.s16        q10, q8,  q15
  51         vbsl            q3,  q13, q14
  52         vbsl            q11, q13, q14
  53         vadd.s16        q2,  q2,  q3
  54         vceq.s16        q9,  q8,  #0
  55         vadd.s16        q10, q10, q11
  56         vbif            q0,  q2,  q1
  57         vbif            q8,  q10, q9
  58         subs            r3,  r3,  #16
  59         vst1.16         {q0},     [r1,:128]!
  60         vst1.16         {q8},     [r1,:128]!
  61         bxle            lr
  62         cmp             r3,  #8
  63         bgt             1b
  64 2:
  65         vld1.16         {d0},     [r0,:64]
  66         vclt.s16        d3,  d0,  #0
  67         vceq.s16        d1,  d0,  #0
  68         vmul.s16        d2,  d0,  d30
  69         vbsl            d3,  d26, d28
  70         vadd.s16        d2,  d2,  d3
  71         vbif            d0,  d2,  d1
  72         vst1.16         {d0},     [r1,:64]
  73         bx              lr
  74 endfunc
  75
  76 function ff_dct_unquantize_h263_intra_neon, export=1
  77         push            {r4-r6,lr}
  78         ldr             r6,  [r0, #AC_PRED]
  79         add             r5,  r0,  #0x2700
  80         cmp             r6,  #0
  81         movne           r12, #63
  82         bne             1f
  83         add             lr,  r0,  #0x2200
  84         add             r12, lr,  #BLOCK_LAST_INDEX-0x2200
  85         add             lr,  lr,  #INTER_SCANTAB_RASTER_END-0x2200
  86         ldr             r12, [r12, r2, lsl #2]
  87         ldrb            r12, [lr, r12]
  88 1:      ldr             r5,  [r5, #H263_AIC-0x2700]
  89         ldrsh           r4,  [r1]
  90         cmp             r5,  #0
  91         mov             r5,  r1
  92         movne           r2,  #0
  93         bne             2f
  94         cmp             r2,  #4
  95         addge           r0,  r0,  #4
  96         sub             r2,  r3,  #1
  97         ldr             r6,  [r0, #Y_DC_SCALE]
  98         orr             r2,  r2,  #1
  99         smulbb          r4,  r4,  r6
 100 2:      lsl             r0,  r3,  #1
 101         add             r3,  r12, #1
 102         bl              ff_dct_unquantize_h263_neon
 103         vmov.16         d0[0], r4
 104         vst1.16         {d0[0]},  [r5]
 105         pop             {r4-r6,pc}
 106 endfunc