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 "libavutil/attributes.h" 22#include "libavutil/mips/cpu.h" 23#include "config.h" 24#include "libavutil/common.h" 25#include "libavcodec/vp9dsp.h" 26#include "vp9dsp_mips.h" 27 28#if HAVE_MSA 29static av_cold void vp9dsp_intrapred_init_msa(VP9DSPContext *dsp, int bpp) 30{ 31 if (bpp == 8) { 32#define init_intra_pred_msa(tx, sz) \ 33 dsp->intra_pred[tx][VERT_PRED] = ff_vert_##sz##_msa; \ 34 dsp->intra_pred[tx][HOR_PRED] = ff_hor_##sz##_msa; \ 35 dsp->intra_pred[tx][DC_PRED] = ff_dc_##sz##_msa; \ 36 dsp->intra_pred[tx][LEFT_DC_PRED] = ff_dc_left_##sz##_msa; \ 37 dsp->intra_pred[tx][TOP_DC_PRED] = ff_dc_top_##sz##_msa; \ 38 dsp->intra_pred[tx][DC_128_PRED] = ff_dc_128_##sz##_msa; \ 39 dsp->intra_pred[tx][DC_127_PRED] = ff_dc_127_##sz##_msa; \ 40 dsp->intra_pred[tx][DC_129_PRED] = ff_dc_129_##sz##_msa; \ 41 dsp->intra_pred[tx][TM_VP8_PRED] = ff_tm_##sz##_msa; \ 42 43 init_intra_pred_msa(TX_16X16, 16x16); 44 init_intra_pred_msa(TX_32X32, 32x32); 45#undef init_intra_pred_msa 46 47#define init_intra_pred_msa(tx, sz) \ 48 dsp->intra_pred[tx][DC_PRED] = ff_dc_##sz##_msa; \ 49 dsp->intra_pred[tx][LEFT_DC_PRED] = ff_dc_left_##sz##_msa; \ 50 dsp->intra_pred[tx][TOP_DC_PRED] = ff_dc_top_##sz##_msa; \ 51 dsp->intra_pred[tx][TM_VP8_PRED] = ff_tm_##sz##_msa; \ 52 53 init_intra_pred_msa(TX_4X4, 4x4); 54 init_intra_pred_msa(TX_8X8, 8x8); 55#undef init_intra_pred_msa 56 } 57} 58 59static av_cold void vp9dsp_itxfm_init_msa(VP9DSPContext *dsp, int bpp) 60{ 61 if (bpp == 8) { 62#define init_itxfm(tx, sz) \ 63 dsp->itxfm_add[tx][DCT_DCT] = ff_idct_idct_##sz##_add_msa; \ 64 dsp->itxfm_add[tx][DCT_ADST] = ff_iadst_idct_##sz##_add_msa; \ 65 dsp->itxfm_add[tx][ADST_DCT] = ff_idct_iadst_##sz##_add_msa; \ 66 dsp->itxfm_add[tx][ADST_ADST] = ff_iadst_iadst_##sz##_add_msa \ 67 68#define init_idct(tx, nm) \ 69 dsp->itxfm_add[tx][DCT_DCT] = \ 70 dsp->itxfm_add[tx][ADST_DCT] = \ 71 dsp->itxfm_add[tx][DCT_ADST] = \ 72 dsp->itxfm_add[tx][ADST_ADST] = nm##_add_msa 73 74 init_itxfm(TX_4X4, 4x4); 75 init_itxfm(TX_8X8, 8x8); 76 init_itxfm(TX_16X16, 16x16); 77 init_idct(TX_32X32, ff_idct_idct_32x32); 78#undef init_itxfm 79#undef init_idct 80 } 81} 82 83static av_cold void vp9dsp_mc_init_msa(VP9DSPContext *dsp, int bpp) 84{ 85 if (bpp == 8) { 86#define init_fpel(idx1, idx2, sz, type) \ 87 dsp->mc[idx1][FILTER_8TAP_SMOOTH ][idx2][0][0] = ff_##type##sz##_msa; \ 88 dsp->mc[idx1][FILTER_8TAP_REGULAR][idx2][0][0] = ff_##type##sz##_msa; \ 89 dsp->mc[idx1][FILTER_8TAP_SHARP ][idx2][0][0] = ff_##type##sz##_msa; \ 90 dsp->mc[idx1][FILTER_BILINEAR ][idx2][0][0] = ff_##type##sz##_msa 91 92#define init_copy_avg(idx, sz) \ 93 init_fpel(idx, 0, sz, copy); \ 94 init_fpel(idx, 1, sz, avg) 95 96#define init_avg(idx, sz) \ 97 init_fpel(idx, 1, sz, avg) 98 99 init_copy_avg(0, 64); 100 init_copy_avg(1, 32); 101 init_copy_avg(2, 16); 102 init_copy_avg(3, 8); 103 init_avg(4, 4); 104 105#undef init_copy_avg 106#undef init_avg 107#undef init_fpel 108 109#define init_subpel1(idx1, idx2, idxh, idxv, sz, dir, type) \ 110 dsp->mc[idx1][FILTER_BILINEAR ][idx2][idxh][idxv] = \ 111 ff_##type##_bilin_##sz##dir##_msa; \ 112 dsp->mc[idx1][FILTER_8TAP_SMOOTH ][idx2][idxh][idxv] = \ 113 ff_##type##_8tap_smooth_##sz##dir##_msa; \ 114 dsp->mc[idx1][FILTER_8TAP_REGULAR][idx2][idxh][idxv] = \ 115 ff_##type##_8tap_regular_##sz##dir##_msa; \ 116 dsp->mc[idx1][FILTER_8TAP_SHARP ][idx2][idxh][idxv] = \ 117 ff_##type##_8tap_sharp_##sz##dir##_msa; 118 119#define init_subpel2(idx, idxh, idxv, dir, type) \ 120 init_subpel1(0, idx, idxh, idxv, 64, dir, type); \ 121 init_subpel1(1, idx, idxh, idxv, 32, dir, type); \ 122 init_subpel1(2, idx, idxh, idxv, 16, dir, type); \ 123 init_subpel1(3, idx, idxh, idxv, 8, dir, type); \ 124 init_subpel1(4, idx, idxh, idxv, 4, dir, type) 125 126#define init_subpel3(idx, type) \ 127 init_subpel2(idx, 1, 1, hv, type); \ 128 init_subpel2(idx, 0, 1, v, type); \ 129 init_subpel2(idx, 1, 0, h, type) 130 131 init_subpel3(0, put); 132 init_subpel3(1, avg); 133 134#undef init_subpel1 135#undef init_subpel2 136#undef init_subpel3 137 } 138} 139 140static av_cold void vp9dsp_loopfilter_init_msa(VP9DSPContext *dsp, int bpp) 141{ 142 if (bpp == 8) { 143 dsp->loop_filter_8[0][0] = ff_loop_filter_h_4_8_msa; 144 dsp->loop_filter_8[0][1] = ff_loop_filter_v_4_8_msa; 145 dsp->loop_filter_8[1][0] = ff_loop_filter_h_8_8_msa; 146 dsp->loop_filter_8[1][1] = ff_loop_filter_v_8_8_msa; 147 dsp->loop_filter_8[2][0] = ff_loop_filter_h_16_8_msa; 148 dsp->loop_filter_8[2][1] = ff_loop_filter_v_16_8_msa; 149 150 dsp->loop_filter_16[0] = ff_loop_filter_h_16_16_msa; 151 dsp->loop_filter_16[1] = ff_loop_filter_v_16_16_msa; 152 153 dsp->loop_filter_mix2[0][0][0] = ff_loop_filter_h_44_16_msa; 154 dsp->loop_filter_mix2[0][0][1] = ff_loop_filter_v_44_16_msa; 155 dsp->loop_filter_mix2[0][1][0] = ff_loop_filter_h_48_16_msa; 156 dsp->loop_filter_mix2[0][1][1] = ff_loop_filter_v_48_16_msa; 157 dsp->loop_filter_mix2[1][0][0] = ff_loop_filter_h_84_16_msa; 158 dsp->loop_filter_mix2[1][0][1] = ff_loop_filter_v_84_16_msa; 159 dsp->loop_filter_mix2[1][1][0] = ff_loop_filter_h_88_16_msa; 160 dsp->loop_filter_mix2[1][1][1] = ff_loop_filter_v_88_16_msa; 161 } 162} 163 164static av_cold void vp9dsp_init_msa(VP9DSPContext *dsp, int bpp) 165{ 166 vp9dsp_intrapred_init_msa(dsp, bpp); 167 vp9dsp_itxfm_init_msa(dsp, bpp); 168 vp9dsp_mc_init_msa(dsp, bpp); 169 vp9dsp_loopfilter_init_msa(dsp, bpp); 170} 171#endif // #if HAVE_MSA 172 173#if HAVE_MMI 174static av_cold void vp9dsp_mc_init_mmi(VP9DSPContext *dsp) 175{ 176#define init_subpel1(idx1, idx2, idxh, idxv, sz, dir, type) \ 177 dsp->mc[idx1][FILTER_8TAP_SMOOTH ][idx2][idxh][idxv] = \ 178 ff_##type##_8tap_smooth_##sz##dir##_mmi; \ 179 dsp->mc[idx1][FILTER_8TAP_REGULAR][idx2][idxh][idxv] = \ 180 ff_##type##_8tap_regular_##sz##dir##_mmi; \ 181 dsp->mc[idx1][FILTER_8TAP_SHARP ][idx2][idxh][idxv] = \ 182 ff_##type##_8tap_sharp_##sz##dir##_mmi; 183 184#define init_subpel2(idx, idxh, idxv, dir, type) \ 185 init_subpel1(0, idx, idxh, idxv, 64, dir, type); \ 186 init_subpel1(1, idx, idxh, idxv, 32, dir, type); \ 187 init_subpel1(2, idx, idxh, idxv, 16, dir, type); \ 188 init_subpel1(3, idx, idxh, idxv, 8, dir, type); \ 189 init_subpel1(4, idx, idxh, idxv, 4, dir, type) 190 191#define init_subpel3(idx, type) \ 192 init_subpel2(idx, 1, 1, hv, type); \ 193 init_subpel2(idx, 0, 1, v, type); \ 194 init_subpel2(idx, 1, 0, h, type) 195 196 init_subpel3(0, put); 197 init_subpel3(1, avg); 198 199#undef init_subpel1 200#undef init_subpel2 201#undef init_subpel3 202} 203 204static av_cold void vp9dsp_init_mmi(VP9DSPContext *dsp, int bpp) 205{ 206 if (bpp == 8) { 207 vp9dsp_mc_init_mmi(dsp); 208 } 209} 210#endif // #if HAVE_MMI 211 212av_cold void ff_vp9dsp_init_mips(VP9DSPContext *dsp, int bpp) 213{ 214#if HAVE_MSA || HAVE_MMI 215 int cpu_flags = av_get_cpu_flags(); 216#endif 217 218#if HAVE_MMI 219 if (have_mmi(cpu_flags)) 220 vp9dsp_init_mmi(dsp, bpp); 221#endif 222 223#if HAVE_MSA 224 if (have_msa(cpu_flags)) 225 vp9dsp_init_msa(dsp, bpp); 226#endif 227} 228