FFmpeg
vp9dsp_init_mips.c
Go to the documentation of this file.
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 #include "config.h"
22 #include "libavutil/common.h"
23 #include "libavcodec/vp9dsp.h"
24 #include "vp9dsp_mips.h"
25 
26 #if HAVE_MSA
27 static av_cold void vp9dsp_intrapred_init_msa(VP9DSPContext *dsp, int bpp)
28 {
29  if (bpp == 8) {
30 #define init_intra_pred_msa(tx, sz) \
31  dsp->intra_pred[tx][VERT_PRED] = ff_vert_##sz##_msa; \
32  dsp->intra_pred[tx][HOR_PRED] = ff_hor_##sz##_msa; \
33  dsp->intra_pred[tx][DC_PRED] = ff_dc_##sz##_msa; \
34  dsp->intra_pred[tx][LEFT_DC_PRED] = ff_dc_left_##sz##_msa; \
35  dsp->intra_pred[tx][TOP_DC_PRED] = ff_dc_top_##sz##_msa; \
36  dsp->intra_pred[tx][DC_128_PRED] = ff_dc_128_##sz##_msa; \
37  dsp->intra_pred[tx][DC_127_PRED] = ff_dc_127_##sz##_msa; \
38  dsp->intra_pred[tx][DC_129_PRED] = ff_dc_129_##sz##_msa; \
39  dsp->intra_pred[tx][TM_VP8_PRED] = ff_tm_##sz##_msa; \
40 
41  init_intra_pred_msa(TX_16X16, 16x16);
42  init_intra_pred_msa(TX_32X32, 32x32);
43 #undef init_intra_pred_msa
44 
45 #define init_intra_pred_msa(tx, sz) \
46  dsp->intra_pred[tx][DC_PRED] = ff_dc_##sz##_msa; \
47  dsp->intra_pred[tx][LEFT_DC_PRED] = ff_dc_left_##sz##_msa; \
48  dsp->intra_pred[tx][TOP_DC_PRED] = ff_dc_top_##sz##_msa; \
49  dsp->intra_pred[tx][TM_VP8_PRED] = ff_tm_##sz##_msa; \
50 
51  init_intra_pred_msa(TX_4X4, 4x4);
52  init_intra_pred_msa(TX_8X8, 8x8);
53 #undef init_intra_pred_msa
54  }
55 }
56 
57 static av_cold void vp9dsp_itxfm_init_msa(VP9DSPContext *dsp, int bpp)
58 {
59  if (bpp == 8) {
60 #define init_itxfm(tx, sz) \
61  dsp->itxfm_add[tx][DCT_DCT] = ff_idct_idct_##sz##_add_msa; \
62  dsp->itxfm_add[tx][DCT_ADST] = ff_iadst_idct_##sz##_add_msa; \
63  dsp->itxfm_add[tx][ADST_DCT] = ff_idct_iadst_##sz##_add_msa; \
64  dsp->itxfm_add[tx][ADST_ADST] = ff_iadst_iadst_##sz##_add_msa \
65 
66 #define init_idct(tx, nm) \
67  dsp->itxfm_add[tx][DCT_DCT] = \
68  dsp->itxfm_add[tx][ADST_DCT] = \
69  dsp->itxfm_add[tx][DCT_ADST] = \
70  dsp->itxfm_add[tx][ADST_ADST] = nm##_add_msa
71 
72  init_itxfm(TX_4X4, 4x4);
73  init_itxfm(TX_8X8, 8x8);
74  init_itxfm(TX_16X16, 16x16);
75  init_idct(TX_32X32, ff_idct_idct_32x32);
76 #undef init_itxfm
77 #undef init_idct
78  }
79 }
80 
81 static av_cold void vp9dsp_mc_init_msa(VP9DSPContext *dsp, int bpp)
82 {
83  if (bpp == 8) {
84 #define init_fpel(idx1, idx2, sz, type) \
85  dsp->mc[idx1][FILTER_8TAP_SMOOTH ][idx2][0][0] = ff_##type##sz##_msa; \
86  dsp->mc[idx1][FILTER_8TAP_REGULAR][idx2][0][0] = ff_##type##sz##_msa; \
87  dsp->mc[idx1][FILTER_8TAP_SHARP ][idx2][0][0] = ff_##type##sz##_msa; \
88  dsp->mc[idx1][FILTER_BILINEAR ][idx2][0][0] = ff_##type##sz##_msa
89 
90 #define init_copy_avg(idx, sz) \
91  init_fpel(idx, 0, sz, copy); \
92  init_fpel(idx, 1, sz, avg)
93 
94 #define init_avg(idx, sz) \
95  init_fpel(idx, 1, sz, avg)
96 
97  init_copy_avg(0, 64);
98  init_copy_avg(1, 32);
99  init_copy_avg(2, 16);
100  init_copy_avg(3, 8);
101  init_avg(4, 4);
102 
103 #undef init_copy_avg
104 #undef init_avg
105 #undef init_fpel
106 
107 #define init_subpel1(idx1, idx2, idxh, idxv, sz, dir, type) \
108  dsp->mc[idx1][FILTER_BILINEAR ][idx2][idxh][idxv] = \
109  ff_##type##_bilin_##sz##dir##_msa; \
110  dsp->mc[idx1][FILTER_8TAP_SMOOTH ][idx2][idxh][idxv] = \
111  ff_##type##_8tap_smooth_##sz##dir##_msa; \
112  dsp->mc[idx1][FILTER_8TAP_REGULAR][idx2][idxh][idxv] = \
113  ff_##type##_8tap_regular_##sz##dir##_msa; \
114  dsp->mc[idx1][FILTER_8TAP_SHARP ][idx2][idxh][idxv] = \
115  ff_##type##_8tap_sharp_##sz##dir##_msa;
116 
117 #define init_subpel2(idx, idxh, idxv, dir, type) \
118  init_subpel1(0, idx, idxh, idxv, 64, dir, type); \
119  init_subpel1(1, idx, idxh, idxv, 32, dir, type); \
120  init_subpel1(2, idx, idxh, idxv, 16, dir, type); \
121  init_subpel1(3, idx, idxh, idxv, 8, dir, type); \
122  init_subpel1(4, idx, idxh, idxv, 4, dir, type)
123 
124 #define init_subpel3(idx, type) \
125  init_subpel2(idx, 1, 1, hv, type); \
126  init_subpel2(idx, 0, 1, v, type); \
127  init_subpel2(idx, 1, 0, h, type)
128 
129  init_subpel3(0, put);
130  init_subpel3(1, avg);
131 
132 #undef init_subpel1
133 #undef init_subpel2
134 #undef init_subpel3
135  }
136 }
137 
138 static av_cold void vp9dsp_loopfilter_init_msa(VP9DSPContext *dsp, int bpp)
139 {
140  if (bpp == 8) {
147 
150 
159  }
160 }
161 
162 static av_cold void vp9dsp_init_msa(VP9DSPContext *dsp, int bpp)
163 {
164  vp9dsp_intrapred_init_msa(dsp, bpp);
165  vp9dsp_itxfm_init_msa(dsp, bpp);
166  vp9dsp_mc_init_msa(dsp, bpp);
167  vp9dsp_loopfilter_init_msa(dsp, bpp);
168 }
169 #endif // #if HAVE_MSA
170 
171 #if HAVE_MMI
172 static av_cold void vp9dsp_mc_init_mmi(VP9DSPContext *dsp)
173 {
174 #define init_subpel1(idx1, idx2, idxh, idxv, sz, dir, type) \
175  dsp->mc[idx1][FILTER_8TAP_SMOOTH ][idx2][idxh][idxv] = \
176  ff_##type##_8tap_smooth_##sz##dir##_mmi; \
177  dsp->mc[idx1][FILTER_8TAP_REGULAR][idx2][idxh][idxv] = \
178  ff_##type##_8tap_regular_##sz##dir##_mmi; \
179  dsp->mc[idx1][FILTER_8TAP_SHARP ][idx2][idxh][idxv] = \
180  ff_##type##_8tap_sharp_##sz##dir##_mmi;
181 
182 #define init_subpel2(idx, idxh, idxv, dir, type) \
183  init_subpel1(0, idx, idxh, idxv, 64, dir, type); \
184  init_subpel1(1, idx, idxh, idxv, 32, dir, type); \
185  init_subpel1(2, idx, idxh, idxv, 16, dir, type); \
186  init_subpel1(3, idx, idxh, idxv, 8, dir, type); \
187  init_subpel1(4, idx, idxh, idxv, 4, dir, type)
188 
189 #define init_subpel3(idx, type) \
190  init_subpel2(idx, 1, 1, hv, type); \
191  init_subpel2(idx, 0, 1, v, type); \
192  init_subpel2(idx, 1, 0, h, type)
193 
194  init_subpel3(0, put);
195  init_subpel3(1, avg);
196 
197 #undef init_subpel1
198 #undef init_subpel2
199 #undef init_subpel3
200 }
201 
202 static av_cold void vp9dsp_init_mmi(VP9DSPContext *dsp, int bpp)
203 {
204  if (bpp == 8) {
205  vp9dsp_mc_init_mmi(dsp);
206  }
207 }
208 #endif // #if HAVE_MMI
209 
211 {
212 #if HAVE_MMI
213  vp9dsp_init_mmi(dsp, bpp);
214 #endif // #if HAVE_MMI
215 #if HAVE_MSA
216  vp9dsp_init_msa(dsp, bpp);
217 #endif // #if HAVE_MSA
218 }
ff_loop_filter_h_88_16_msa
void ff_loop_filter_h_88_16_msa(uint8_t *src, ptrdiff_t pitch, int32_t b_limit_ptr, int32_t limit_ptr, int32_t thresh_ptr)
Definition: vp9_lpf_msa.c:1349
ff_loop_filter_v_44_16_msa
void ff_loop_filter_v_44_16_msa(uint8_t *src, ptrdiff_t pitch, int32_t b_limit_ptr, int32_t limit_ptr, int32_t thresh_ptr)
Definition: vp9_lpf_msa.c:225
ff_loop_filter_h_8_8_msa
void ff_loop_filter_h_8_8_msa(uint8_t *src, ptrdiff_t pitch, int32_t b_limit_ptr, int32_t limit_ptr, int32_t thresh_ptr)
Definition: vp9_lpf_msa.c:1271
ff_loop_filter_v_4_8_msa
void ff_loop_filter_v_4_8_msa(uint8_t *src, ptrdiff_t pitch, int32_t b_limit_ptr, int32_t limit_ptr, int32_t thresh_ptr)
Definition: vp9_lpf_msa.c:196
ff_loop_filter_h_48_16_msa
void ff_loop_filter_h_48_16_msa(uint8_t *src, ptrdiff_t pitch, int32_t b_limit_ptr, int32_t limit_ptr, int32_t thresh_ptr)
Definition: vp9_lpf_msa.c:1565
VP9DSPContext::loop_filter_8
void(* loop_filter_8[3][2])(uint8_t *dst, ptrdiff_t stride, int mb_lim, int lim, int hev_thr)
Definition: vp9dsp.h:80
VP9DSPContext
Definition: vp9dsp.h:39
VP9DSPContext::loop_filter_mix2
void(* loop_filter_mix2[2][2][2])(uint8_t *dst, ptrdiff_t stride, int mb_lim, int lim, int hev_thr)
Definition: vp9dsp.h:102
av_cold
#define av_cold
Definition: attributes.h:90
init_avg
#define init_avg(idx, sz, suffix)
ff_loop_filter_v_48_16_msa
void ff_loop_filter_v_48_16_msa(uint8_t *src, ptrdiff_t pitch, int32_t b_limit_ptr, int32_t limit_ptr, int32_t thresh_ptr)
Definition: vp9_lpf_msa.c:479
vp9dsp_mips.h
ff_loop_filter_h_84_16_msa
void ff_loop_filter_h_84_16_msa(uint8_t *src, ptrdiff_t pitch, int32_t b_limit_ptr, int32_t limit_ptr, int32_t thresh_ptr)
Definition: vp9_lpf_msa.c:1462
TX_8X8
@ TX_8X8
Definition: vp9.h:29
ff_loop_filter_v_16_16_msa
void ff_loop_filter_v_16_16_msa(uint8_t *src, ptrdiff_t pitch, int32_t b_limit_ptr, int32_t limit_ptr, int32_t thresh_ptr)
Definition: vp9_lpf_msa.c:958
ff_loop_filter_h_4_8_msa
void ff_loop_filter_h_4_8_msa(uint8_t *src, ptrdiff_t pitch, int32_t b_limit_ptr, int32_t limit_ptr, int32_t thresh_ptr)
Definition: vp9_lpf_msa.c:1198
TX_16X16
@ TX_16X16
Definition: vp9.h:30
init_idct
#define init_idct(tx, nm, bpp)
vp9dsp.h
avg
#define avg(a, b, c, d)
Definition: colorspacedsp_template.c:28
TX_4X4
@ TX_4X4
Definition: vp9.h:28
ff_loop_filter_h_16_8_msa
void ff_loop_filter_h_16_8_msa(uint8_t *src, ptrdiff_t pitch, int32_t b_limit_ptr, int32_t limit_ptr, int32_t thresh_ptr)
Definition: vp9_lpf_msa.c:2066
ff_loop_filter_v_84_16_msa
void ff_loop_filter_v_84_16_msa(uint8_t *src, ptrdiff_t pitch, int32_t b_limit_ptr, int32_t limit_ptr, int32_t thresh_ptr)
Definition: vp9_lpf_msa.c:408
common.h
ff_loop_filter_v_16_8_msa
void ff_loop_filter_v_16_8_msa(uint8_t *src, ptrdiff_t pitch, int32_t b_limit_ptr, int32_t limit_ptr, int32_t thresh_ptr)
Definition: vp9_lpf_msa.c:974
ff_loop_filter_h_44_16_msa
void ff_loop_filter_h_44_16_msa(uint8_t *src, ptrdiff_t pitch, int32_t b_limit_ptr, int32_t limit_ptr, int32_t thresh_ptr)
Definition: vp9_lpf_msa.c:1225
init_itxfm
#define init_itxfm(tx, sz, bpp)
ff_vp9dsp_init_mips
av_cold void ff_vp9dsp_init_mips(VP9DSPContext *dsp, int bpp)
Definition: vp9dsp_init_mips.c:210
init_copy_avg
#define init_copy_avg(idx, sz1, sz2)
ff_loop_filter_v_8_8_msa
void ff_loop_filter_v_8_8_msa(uint8_t *src, ptrdiff_t pitch, int32_t b_limit_ptr, int32_t limit_ptr, int32_t thresh_ptr)
Definition: vp9_lpf_msa.c:255
ff_loop_filter_h_16_16_msa
void ff_loop_filter_h_16_16_msa(uint8_t *src, ptrdiff_t pitch, int32_t b_limit_ptr, int32_t limit_ptr, int32_t thresh_ptr)
Definition: vp9_lpf_msa.c:2502
TX_32X32
@ TX_32X32
Definition: vp9.h:31
ff_loop_filter_v_88_16_msa
void ff_loop_filter_v_88_16_msa(uint8_t *src, ptrdiff_t pitch, int32_t b_limit_ptr, int32_t limit_ptr, int32_t thresh_ptr)
Definition: vp9_lpf_msa.c:329
init_subpel3
#define init_subpel3(idx, type)
VP9DSPContext::loop_filter_16
void(* loop_filter_16[2])(uint8_t *dst, ptrdiff_t stride, int mb_lim, int lim, int hev_thr)
Definition: vp9dsp.h:88