1 /*
2  * Copyright (c) 2015 Shivraj Patil (Shivraj.Patil@imgtec.com)
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 #ifndef AVCODEC_MIPS_VP9DSP_MIPS_H
22 #define AVCODEC_MIPS_VP9DSP_MIPS_H
23 
24 #include <stddef.h>
25 #include <stdint.h>
26 
27 #define VP9_8TAP_MIPS_MSA_FUNC(SIZE, type, type_idx)                         \
28 void ff_put_8tap_##type##_##SIZE##h_msa(uint8_t *dst, ptrdiff_t dststride,   \
29                                         const uint8_t *src,                  \
30                                         ptrdiff_t srcstride,                 \
31                                         int h, int mx, int my);              \
32                                                                              \
33 void ff_put_8tap_##type##_##SIZE##v_msa(uint8_t *dst, ptrdiff_t dststride,   \
34                                         const uint8_t *src,                  \
35                                         ptrdiff_t srcstride,                 \
36                                         int h, int mx, int my);              \
37                                                                              \
38 void ff_put_8tap_##type##_##SIZE##hv_msa(uint8_t *dst, ptrdiff_t dststride,  \
39                                          const uint8_t *src,                 \
40                                          ptrdiff_t srcstride,                \
41                                          int h, int mx, int my);             \
42                                                                              \
43 void ff_avg_8tap_##type##_##SIZE##h_msa(uint8_t *dst, ptrdiff_t dststride,   \
44                                         const uint8_t *src,                  \
45                                         ptrdiff_t srcstride,                 \
46                                         int h, int mx, int my);              \
47                                                                              \
48 void ff_avg_8tap_##type##_##SIZE##v_msa(uint8_t *dst, ptrdiff_t dststride,   \
49                                         const uint8_t *src,                  \
50                                         ptrdiff_t srcstride,                 \
51                                         int h, int mx, int my);              \
52                                                                              \
53 void ff_avg_8tap_##type##_##SIZE##hv_msa(uint8_t *dst, ptrdiff_t dststride,  \
54                                          const uint8_t *src,                 \
55                                          ptrdiff_t srcstride,                \
56                                          int h, int mx, int my);
57 
58 #define VP9_BILINEAR_MIPS_MSA_FUNC(SIZE)                                   \
59 void ff_put_bilin_##SIZE##h_msa(uint8_t *dst, ptrdiff_t dststride,         \
60                                 const uint8_t *src, ptrdiff_t srcstride,   \
61                                 int h, int mx, int my);                    \
62                                                                            \
63 void ff_put_bilin_##SIZE##v_msa(uint8_t *dst, ptrdiff_t dststride,         \
64                                 const uint8_t *src, ptrdiff_t srcstride,   \
65                                 int h, int mx, int my);                    \
66                                                                            \
67 void ff_put_bilin_##SIZE##hv_msa(uint8_t *dst, ptrdiff_t dststride,        \
68                                  const uint8_t *src, ptrdiff_t srcstride,  \
69                                  int h, int mx, int my);                   \
70                                                                            \
71 void ff_avg_bilin_##SIZE##h_msa(uint8_t *dst, ptrdiff_t dststride,         \
72                                 const uint8_t *src, ptrdiff_t srcstride,   \
73                                 int h, int mx, int my);                    \
74                                                                            \
75 void ff_avg_bilin_##SIZE##v_msa(uint8_t *dst, ptrdiff_t dststride,         \
76                                 const uint8_t *src, ptrdiff_t srcstride,   \
77                                 int h, int mx, int my);                    \
78                                                                            \
79 void ff_avg_bilin_##SIZE##hv_msa(uint8_t *dst, ptrdiff_t dststride,        \
80                                  const uint8_t *src, ptrdiff_t srcstride,  \
81                                  int h, int mx, int my);
82 
83 #define VP9_COPY_AVG_MIPS_MSA_FUNC(SIZE)                           \
84 void ff_copy##SIZE##_msa(uint8_t *dst, ptrdiff_t dststride,        \
85                          const uint8_t *src, ptrdiff_t srcstride,  \
86                          int h, int mx, int my);                   \
87                                                                    \
88 void ff_avg##SIZE##_msa(uint8_t *dst, ptrdiff_t dststride,         \
89                         const uint8_t *src, ptrdiff_t srcstride,   \
90                         int h, int mx, int my);
91 
92 VP9_8TAP_MIPS_MSA_FUNC(64, regular, FILTER_8TAP_REGULAR);
93 VP9_8TAP_MIPS_MSA_FUNC(32, regular, FILTER_8TAP_REGULAR);
94 VP9_8TAP_MIPS_MSA_FUNC(16, regular, FILTER_8TAP_REGULAR);
95 VP9_8TAP_MIPS_MSA_FUNC(8, regular, FILTER_8TAP_REGULAR);
96 VP9_8TAP_MIPS_MSA_FUNC(4, regular, FILTER_8TAP_REGULAR);
97 
98 VP9_8TAP_MIPS_MSA_FUNC(64, sharp, FILTER_8TAP_SHARP);
99 VP9_8TAP_MIPS_MSA_FUNC(32, sharp, FILTER_8TAP_SHARP);
100 VP9_8TAP_MIPS_MSA_FUNC(16, sharp, FILTER_8TAP_SHARP);
101 VP9_8TAP_MIPS_MSA_FUNC(8, sharp, FILTER_8TAP_SHARP);
102 VP9_8TAP_MIPS_MSA_FUNC(4, sharp, FILTER_8TAP_SHARP);
103 
104 VP9_8TAP_MIPS_MSA_FUNC(64, smooth, FILTER_8TAP_SMOOTH);
105 VP9_8TAP_MIPS_MSA_FUNC(32, smooth, FILTER_8TAP_SMOOTH);
106 VP9_8TAP_MIPS_MSA_FUNC(16, smooth, FILTER_8TAP_SMOOTH);
107 VP9_8TAP_MIPS_MSA_FUNC(8, smooth, FILTER_8TAP_SMOOTH);
108 VP9_8TAP_MIPS_MSA_FUNC(4, smooth, FILTER_8TAP_SMOOTH);
109 
110 VP9_BILINEAR_MIPS_MSA_FUNC(64);
111 VP9_BILINEAR_MIPS_MSA_FUNC(32);
112 VP9_BILINEAR_MIPS_MSA_FUNC(16);
113 VP9_BILINEAR_MIPS_MSA_FUNC(8);
114 VP9_BILINEAR_MIPS_MSA_FUNC(4);
115 
116 VP9_COPY_AVG_MIPS_MSA_FUNC(64);
117 VP9_COPY_AVG_MIPS_MSA_FUNC(32);
118 VP9_COPY_AVG_MIPS_MSA_FUNC(16);
119 VP9_COPY_AVG_MIPS_MSA_FUNC(8);
120 VP9_COPY_AVG_MIPS_MSA_FUNC(4);
121 
122 #undef VP9_8TAP_MIPS_MSA_FUNC
123 #undef VP9_BILINEAR_MIPS_MSA_FUNC
124 #undef VP9_COPY_AVG_MIPS_MSA_FUNC
125 
126 void ff_loop_filter_h_4_8_msa(uint8_t *dst, ptrdiff_t stride, int32_t e,
127                               int32_t i, int32_t h);
128 void ff_loop_filter_h_8_8_msa(uint8_t *dst, ptrdiff_t stride, int32_t e,
129                               int32_t i, int32_t h);
130 void ff_loop_filter_h_16_8_msa(uint8_t *dst, ptrdiff_t stride, int32_t e,
131                                int32_t i, int32_t h);
132 void ff_loop_filter_v_4_8_msa(uint8_t *dst, ptrdiff_t stride, int32_t e,
133                               int32_t i, int32_t h);
134 void ff_loop_filter_v_8_8_msa(uint8_t *dst, ptrdiff_t stride, int32_t e,
135                               int32_t i, int32_t h);
136 void ff_loop_filter_v_16_8_msa(uint8_t *dst, ptrdiff_t stride, int32_t e,
137                                int32_t i, int32_t h);
138 void ff_loop_filter_h_44_16_msa(uint8_t *dst, ptrdiff_t stride, int32_t e,
139                                 int32_t i, int32_t h);
140 void ff_loop_filter_h_88_16_msa(uint8_t *dst, ptrdiff_t stride, int32_t e,
141                                 int32_t i, int32_t h);
142 void ff_loop_filter_h_16_16_msa(uint8_t *dst, ptrdiff_t stride, int32_t e,
143                                 int32_t i, int32_t h);
144 void ff_loop_filter_v_44_16_msa(uint8_t *dst, ptrdiff_t stride, int32_t e,
145                                 int32_t i, int32_t h);
146 void ff_loop_filter_v_88_16_msa(uint8_t *dst, ptrdiff_t stride, int32_t e,
147                                 int32_t i, int32_t h);
148 void ff_loop_filter_v_16_16_msa(uint8_t *dst, ptrdiff_t stride, int32_t e,
149                                 int32_t i, int32_t h);
150 void ff_loop_filter_h_48_16_msa(uint8_t *dst, ptrdiff_t stride, int32_t e,
151                                 int32_t i, int32_t h);
152 void ff_loop_filter_h_84_16_msa(uint8_t *dst, ptrdiff_t stride, int32_t e,
153                                 int32_t i, int32_t h);
154 void ff_loop_filter_v_48_16_msa(uint8_t *dst, ptrdiff_t stride, int32_t e,
155                                 int32_t i, int32_t h);
156 void ff_loop_filter_v_84_16_msa(uint8_t *dst, ptrdiff_t stride, int32_t e,
157                                 int32_t i, int32_t h);
158 void ff_idct_idct_4x4_add_msa(uint8_t *dst, ptrdiff_t stride,
159                               int16_t *block, int eob);
160 void ff_idct_idct_8x8_add_msa(uint8_t *dst, ptrdiff_t stride,
161                               int16_t *block, int eob);
162 void ff_idct_idct_16x16_add_msa(uint8_t *dst, ptrdiff_t stride,
163                                 int16_t *block, int eob);
164 void ff_idct_idct_32x32_add_msa(uint8_t *dst, ptrdiff_t stride,
165                                 int16_t *block, int eob);
166 void ff_iadst_iadst_4x4_add_msa(uint8_t *dst, ptrdiff_t stride,
167                                 int16_t *block, int eob);
168 void ff_iadst_iadst_8x8_add_msa(uint8_t *dst, ptrdiff_t stride,
169                                 int16_t *block, int eob);
170 void ff_iadst_iadst_16x16_add_msa(uint8_t *dst, ptrdiff_t stride,
171                                   int16_t *block, int eob);
172 void ff_iadst_idct_4x4_add_msa(uint8_t *dst, ptrdiff_t stride,
173                                int16_t *block, int eob);
174 void ff_iadst_idct_8x8_add_msa(uint8_t *dst, ptrdiff_t stride,
175                                int16_t *block, int eob);
176 void ff_iadst_idct_16x16_add_msa(uint8_t *dst, ptrdiff_t stride,
177                                  int16_t *block, int eob);
178 void ff_idct_iadst_4x4_add_msa(uint8_t *pu8Dest, ptrdiff_t stride,
179                                int16_t *block, int eob);
180 void ff_idct_iadst_8x8_add_msa(uint8_t *pu8Dest, ptrdiff_t stride,
181                                int16_t *block, int eob);
182 void ff_idct_iadst_16x16_add_msa(uint8_t *pu8Dest, ptrdiff_t stride,
183                                  int16_t *block, int eob);
184 void ff_iwht_iwht_4x4_add_msa(uint8_t *dst, ptrdiff_t stride,
185                               int16_t *block, int eob);
186 
187 void ff_vert_16x16_msa(uint8_t *dst, ptrdiff_t stride, const uint8_t *left,
188                        const uint8_t *top);
189 void ff_vert_32x32_msa(uint8_t *dst, ptrdiff_t stride, const uint8_t *left,
190                        const uint8_t *top);
191 void ff_hor_16x16_msa(uint8_t *dst, ptrdiff_t stride, const uint8_t *left,
192                       const uint8_t *top);
193 void ff_hor_32x32_msa(uint8_t *dst, ptrdiff_t stride, const uint8_t *left,
194                       const uint8_t *top);
195 void ff_dc_4x4_msa(uint8_t *dst, ptrdiff_t stride, const uint8_t *left,
196                    const uint8_t *top);
197 void ff_dc_8x8_msa(uint8_t *dst, ptrdiff_t stride, const uint8_t *left,
198                    const uint8_t *top);
199 void ff_dc_16x16_msa(uint8_t *dst, ptrdiff_t stride, const uint8_t *left,
200                      const uint8_t *top);
201 void ff_dc_32x32_msa(uint8_t *dst, ptrdiff_t stride, const uint8_t *left,
202                      const uint8_t *top);
203 void ff_dc_left_4x4_msa(uint8_t *dst, ptrdiff_t stride, const uint8_t *left,
204                         const uint8_t *top);
205 void ff_dc_left_8x8_msa(uint8_t *dst, ptrdiff_t stride, const uint8_t *left,
206                         const uint8_t *top);
207 void ff_dc_left_16x16_msa(uint8_t *dst, ptrdiff_t stride,
208                           const uint8_t *left, const uint8_t *top);
209 void ff_dc_left_32x32_msa(uint8_t *dst, ptrdiff_t stride,
210                           const uint8_t *left, const uint8_t *top);
211 void ff_dc_top_4x4_msa(uint8_t *dst, ptrdiff_t stride, const uint8_t *left,
212                        const uint8_t *top);
213 void ff_dc_top_8x8_msa(uint8_t *dst, ptrdiff_t stride, const uint8_t *left,
214                        const uint8_t *top);
215 void ff_dc_top_16x16_msa(uint8_t *dst, ptrdiff_t stride,
216                          const uint8_t *left, const uint8_t *top);
217 void ff_dc_top_32x32_msa(uint8_t *dst, ptrdiff_t stride,
218                          const uint8_t *left, const uint8_t *top);
219 void ff_dc_128_16x16_msa(uint8_t *dst, ptrdiff_t stride,
220                          const uint8_t *left, const uint8_t *top);
221 void ff_dc_128_32x32_msa(uint8_t *dst, ptrdiff_t stride,
222                          const uint8_t *left, const uint8_t *top);
223 void ff_dc_127_16x16_msa(uint8_t *dst, ptrdiff_t stride,
224                          const uint8_t *left, const uint8_t *top);
225 void ff_dc_127_32x32_msa(uint8_t *dst, ptrdiff_t stride,
226                          const uint8_t *left, const uint8_t *top);
227 void ff_dc_129_16x16_msa(uint8_t *dst, ptrdiff_t stride,
228                          const uint8_t *left, const uint8_t *top);
229 void ff_dc_129_32x32_msa(uint8_t *dst, ptrdiff_t stride,
230                          const uint8_t *left, const uint8_t *top);
231 void ff_tm_4x4_msa(uint8_t *dst, ptrdiff_t stride, const uint8_t *left,
232                    const uint8_t *top);
233 void ff_tm_8x8_msa(uint8_t *dst, ptrdiff_t stride, const uint8_t *left,
234                    const uint8_t *top);
235 void ff_tm_16x16_msa(uint8_t *dst, ptrdiff_t stride, const uint8_t *left,
236                      const uint8_t *top);
237 void ff_tm_32x32_msa(uint8_t *dst, ptrdiff_t stride, const uint8_t *left,
238                      const uint8_t *top);
239 
240 #define VP9_8TAP_MIPS_MMI_FUNC(SIZE, type, type_idx)                         \
241 void ff_put_8tap_##type##_##SIZE##h_mmi(uint8_t *dst, ptrdiff_t dststride,   \
242                                         const uint8_t *src,                  \
243                                         ptrdiff_t srcstride,                 \
244                                         int h, int mx, int my);              \
245                                                                              \
246 void ff_put_8tap_##type##_##SIZE##v_mmi(uint8_t *dst, ptrdiff_t dststride,   \
247                                         const uint8_t *src,                  \
248                                         ptrdiff_t srcstride,                 \
249                                         int h, int mx, int my);              \
250                                                                              \
251 void ff_put_8tap_##type##_##SIZE##hv_mmi(uint8_t *dst, ptrdiff_t dststride,  \
252                                          const uint8_t *src,                 \
253                                          ptrdiff_t srcstride,                \
254                                          int h, int mx, int my);             \
255                                                                              \
256 void ff_avg_8tap_##type##_##SIZE##h_mmi(uint8_t *dst, ptrdiff_t dststride,   \
257                                         const uint8_t *src,                  \
258                                         ptrdiff_t srcstride,                 \
259                                         int h, int mx, int my);              \
260                                                                              \
261 void ff_avg_8tap_##type##_##SIZE##v_mmi(uint8_t *dst, ptrdiff_t dststride,   \
262                                         const uint8_t *src,                  \
263                                         ptrdiff_t srcstride,                 \
264                                         int h, int mx, int my);              \
265                                                                              \
266 void ff_avg_8tap_##type##_##SIZE##hv_mmi(uint8_t *dst, ptrdiff_t dststride,  \
267                                          const uint8_t *src,                 \
268                                          ptrdiff_t srcstride,                \
269                                          int h, int mx, int my);
270 
271 VP9_8TAP_MIPS_MMI_FUNC(64, regular, FILTER_8TAP_REGULAR);
272 VP9_8TAP_MIPS_MMI_FUNC(32, regular, FILTER_8TAP_REGULAR);
273 VP9_8TAP_MIPS_MMI_FUNC(16, regular, FILTER_8TAP_REGULAR);
274 VP9_8TAP_MIPS_MMI_FUNC(8, regular, FILTER_8TAP_REGULAR);
275 VP9_8TAP_MIPS_MMI_FUNC(4, regular, FILTER_8TAP_REGULAR);
276 
277 VP9_8TAP_MIPS_MMI_FUNC(64, sharp, FILTER_8TAP_SHARP);
278 VP9_8TAP_MIPS_MMI_FUNC(32, sharp, FILTER_8TAP_SHARP);
279 VP9_8TAP_MIPS_MMI_FUNC(16, sharp, FILTER_8TAP_SHARP);
280 VP9_8TAP_MIPS_MMI_FUNC(8, sharp, FILTER_8TAP_SHARP);
281 VP9_8TAP_MIPS_MMI_FUNC(4, sharp, FILTER_8TAP_SHARP);
282 
283 VP9_8TAP_MIPS_MMI_FUNC(64, smooth, FILTER_8TAP_SMOOTH);
284 VP9_8TAP_MIPS_MMI_FUNC(32, smooth, FILTER_8TAP_SMOOTH);
285 VP9_8TAP_MIPS_MMI_FUNC(16, smooth, FILTER_8TAP_SMOOTH);
286 VP9_8TAP_MIPS_MMI_FUNC(8, smooth, FILTER_8TAP_SMOOTH);
287 VP9_8TAP_MIPS_MMI_FUNC(4, smooth, FILTER_8TAP_SMOOTH);
288 #undef VP9_8TAP_MIPS_MMI_FUNC
289 
290 #endif  // #ifndef AVCODEC_MIPS_VP9DSP_MIPS_H
291