• Home
  • Line#
  • Scopes#
  • Navigate#
  • Raw
  • Download
1 /*
2  * This file is part of FFmpeg.
3  *
4  * FFmpeg is free software; you can redistribute it and/or
5  * modify it under the terms of the GNU Lesser General Public
6  * License as published by the Free Software Foundation; either
7  * version 2.1 of the License, or (at your option) any later version.
8  *
9  * FFmpeg is distributed in the hope that it will be useful,
10  * but WITHOUT ANY WARRANTY; without even the implied warranty of
11  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
12  * Lesser General Public License for more details.
13  *
14  * You should have received a copy of the GNU Lesser General Public
15  * License along with FFmpeg; if not, write to the Free Software
16  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
17  */
18 
19 #include <stdint.h>
20 
21 #include "libavutil/attributes.h"
22 #include "libavcodec/vc1dsp.h"
23 #include "vc1dsp.h"
24 
25 void ff_vc1_inv_trans_8x8_neon(int16_t *block);
26 void ff_vc1_inv_trans_4x8_neon(uint8_t *dest, ptrdiff_t stride, int16_t *block);
27 void ff_vc1_inv_trans_8x4_neon(uint8_t *dest, ptrdiff_t stride, int16_t *block);
28 void ff_vc1_inv_trans_4x4_neon(uint8_t *dest, ptrdiff_t stride, int16_t *block);
29 
30 void ff_vc1_inv_trans_8x8_dc_neon(uint8_t *dest, ptrdiff_t stride, int16_t *block);
31 void ff_vc1_inv_trans_4x8_dc_neon(uint8_t *dest, ptrdiff_t stride, int16_t *block);
32 void ff_vc1_inv_trans_8x4_dc_neon(uint8_t *dest, ptrdiff_t stride, int16_t *block);
33 void ff_vc1_inv_trans_4x4_dc_neon(uint8_t *dest, ptrdiff_t stride, int16_t *block);
34 
35 void ff_put_pixels8x8_neon(uint8_t *block, const uint8_t *pixels,
36                            ptrdiff_t line_size, int rnd);
37 
38 #define DECL_PUT(X, Y) \
39 void ff_put_vc1_mspel_mc##X##Y##_neon(uint8_t *dst, const uint8_t *src, \
40                                       ptrdiff_t stride, int rnd); \
41 static void ff_put_vc1_mspel_mc##X##Y##_16_neon(uint8_t *dst, const uint8_t *src, \
42                                          ptrdiff_t stride, int rnd) \
43 { \
44   ff_put_vc1_mspel_mc##X##Y##_neon(dst+0, src+0, stride, rnd); \
45   ff_put_vc1_mspel_mc##X##Y##_neon(dst+8, src+8, stride, rnd); \
46   dst += 8*stride; src += 8*stride; \
47   ff_put_vc1_mspel_mc##X##Y##_neon(dst+0, src+0, stride, rnd); \
48   ff_put_vc1_mspel_mc##X##Y##_neon(dst+8, src+8, stride, rnd); \
49 }
50 
51 DECL_PUT(1, 0)
52 DECL_PUT(2, 0)
53 DECL_PUT(3, 0)
54 
55 DECL_PUT(0, 1)
56 DECL_PUT(0, 2)
57 DECL_PUT(0, 3)
58 
59 DECL_PUT(1, 1)
60 DECL_PUT(1, 2)
61 DECL_PUT(1, 3)
62 
63 DECL_PUT(2, 1)
64 DECL_PUT(2, 2)
65 DECL_PUT(2, 3)
66 
67 DECL_PUT(3, 1)
68 DECL_PUT(3, 2)
69 DECL_PUT(3, 3)
70 
71 void ff_put_vc1_chroma_mc8_neon(uint8_t *dst, uint8_t *src, ptrdiff_t stride,
72                                 int h, int x, int y);
73 void ff_avg_vc1_chroma_mc8_neon(uint8_t *dst, uint8_t *src, ptrdiff_t stride,
74                                 int h, int x, int y);
75 void ff_put_vc1_chroma_mc4_neon(uint8_t *dst, uint8_t *src, ptrdiff_t stride,
76                                 int h, int x, int y);
77 void ff_avg_vc1_chroma_mc4_neon(uint8_t *dst, uint8_t *src, ptrdiff_t stride,
78                                 int h, int x, int y);
79 
80 #define FN_ASSIGN(X, Y) \
81     dsp->put_vc1_mspel_pixels_tab[0][X+4*Y] = ff_put_vc1_mspel_mc##X##Y##_16_neon; \
82     dsp->put_vc1_mspel_pixels_tab[1][X+4*Y] = ff_put_vc1_mspel_mc##X##Y##_neon
83 
ff_vc1dsp_init_neon(VC1DSPContext * dsp)84 av_cold void ff_vc1dsp_init_neon(VC1DSPContext *dsp)
85 {
86     dsp->vc1_inv_trans_8x8 = ff_vc1_inv_trans_8x8_neon;
87     dsp->vc1_inv_trans_4x8 = ff_vc1_inv_trans_4x8_neon;
88     dsp->vc1_inv_trans_8x4 = ff_vc1_inv_trans_8x4_neon;
89     dsp->vc1_inv_trans_4x4 = ff_vc1_inv_trans_4x4_neon;
90     dsp->vc1_inv_trans_8x8_dc = ff_vc1_inv_trans_8x8_dc_neon;
91     dsp->vc1_inv_trans_4x8_dc = ff_vc1_inv_trans_4x8_dc_neon;
92     dsp->vc1_inv_trans_8x4_dc = ff_vc1_inv_trans_8x4_dc_neon;
93     dsp->vc1_inv_trans_4x4_dc = ff_vc1_inv_trans_4x4_dc_neon;
94 
95     dsp->put_vc1_mspel_pixels_tab[1][ 0] = ff_put_pixels8x8_neon;
96     FN_ASSIGN(1, 0);
97     FN_ASSIGN(2, 0);
98     FN_ASSIGN(3, 0);
99 
100     FN_ASSIGN(0, 1);
101     FN_ASSIGN(1, 1);
102     FN_ASSIGN(2, 1);
103     FN_ASSIGN(3, 1);
104 
105     FN_ASSIGN(0, 2);
106     FN_ASSIGN(1, 2);
107     FN_ASSIGN(2, 2);
108     FN_ASSIGN(3, 2);
109 
110     FN_ASSIGN(0, 3);
111     FN_ASSIGN(1, 3);
112     FN_ASSIGN(2, 3);
113     FN_ASSIGN(3, 3);
114 
115     dsp->put_no_rnd_vc1_chroma_pixels_tab[0] = ff_put_vc1_chroma_mc8_neon;
116     dsp->avg_no_rnd_vc1_chroma_pixels_tab[0] = ff_avg_vc1_chroma_mc8_neon;
117     dsp->put_no_rnd_vc1_chroma_pixels_tab[1] = ff_put_vc1_chroma_mc4_neon;
118     dsp->avg_no_rnd_vc1_chroma_pixels_tab[1] = ff_avg_vc1_chroma_mc4_neon;
119 }
120