vp8dsp_init_arm.c 8.4 KB
Newer Older
1
/**
2
 * This file is part of Libav.
3
 *
4
 * Libav is free software; you can redistribute it and/or
5 6 7 8
 * modify it under the terms of the GNU Lesser General Public
 * License as published by the Free Software Foundation; either
 * version 2.1 of the License, or (at your option) any later version.
 *
9
 * Libav is distributed in the hope that it will be useful,
10 11 12 13 14
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
 * Lesser General Public License for more details.
 *
 * You should have received a copy of the GNU Lesser General Public
15
 * License along with Libav; if not, write to the Free Software
16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163
 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
 */

#include <stdint.h>
#include "libavcodec/vp8dsp.h"

void ff_vp8_luma_dc_wht_neon(DCTELEM block[4][4][16], DCTELEM dc[16]);
void ff_vp8_luma_dc_wht_dc_neon(DCTELEM block[4][4][16], DCTELEM dc[16]);

void ff_vp8_idct_add_neon(uint8_t *dst, DCTELEM block[16], int stride);
void ff_vp8_idct_dc_add_neon(uint8_t *dst, DCTELEM block[16], int stride);
void ff_vp8_idct_dc_add4y_neon(uint8_t *dst, DCTELEM block[4][16], int stride);
void ff_vp8_idct_dc_add4uv_neon(uint8_t *dst, DCTELEM block[4][16], int stride);

void ff_vp8_v_loop_filter16_neon(uint8_t *dst, int stride,
                                 int flim_E, int flim_I, int hev_thresh);
void ff_vp8_h_loop_filter16_neon(uint8_t *dst, int stride,
                                 int flim_E, int flim_I, int hev_thresh);
void ff_vp8_v_loop_filter8uv_neon(uint8_t *dstU, uint8_t *dstV, int stride,
                                  int flim_E, int flim_I, int hev_thresh);
void ff_vp8_h_loop_filter8uv_neon(uint8_t *dstU, uint8_t *dstV, int stride,
                                  int flim_E, int flim_I, int hev_thresh);

void ff_vp8_v_loop_filter16_inner_neon(uint8_t *dst, int stride,
                                       int flim_E, int flim_I, int hev_thresh);
void ff_vp8_h_loop_filter16_inner_neon(uint8_t *dst, int stride,
                                       int flim_E, int flim_I, int hev_thresh);
void ff_vp8_v_loop_filter8uv_inner_neon(uint8_t *dstU, uint8_t *dstV,
                                        int stride, int flim_E, int flim_I,
                                        int hev_thresh);
void ff_vp8_h_loop_filter8uv_inner_neon(uint8_t *dstU, uint8_t *dstV,
                                        int stride, int flim_E, int flim_I,
                                        int hev_thresh);

void ff_vp8_v_loop_filter16_simple_neon(uint8_t *dst, int stride, int flim);
void ff_vp8_h_loop_filter16_simple_neon(uint8_t *dst, int stride, int flim);


#define VP8_MC(n)                                                       \
    void ff_put_vp8_##n##_neon(uint8_t *dst, int dststride,             \
                               uint8_t *src, int srcstride,             \
                               int h, int x, int y)

#define VP8_EPEL(w)                             \
    VP8_MC(pixels ## w);                        \
    VP8_MC(epel ## w ## _h4);                   \
    VP8_MC(epel ## w ## _h6);                   \
    VP8_MC(epel ## w ## _v4);                   \
    VP8_MC(epel ## w ## _h4v4);                 \
    VP8_MC(epel ## w ## _h6v4);                 \
    VP8_MC(epel ## w ## _v6);                   \
    VP8_MC(epel ## w ## _h4v6);                 \
    VP8_MC(epel ## w ## _h6v6)

VP8_EPEL(16);
VP8_EPEL(8);
VP8_EPEL(4);

VP8_MC(bilin16_h);
VP8_MC(bilin16_v);
VP8_MC(bilin16_hv);
VP8_MC(bilin8_h);
VP8_MC(bilin8_v);
VP8_MC(bilin8_hv);
VP8_MC(bilin4_h);
VP8_MC(bilin4_v);
VP8_MC(bilin4_hv);

av_cold void ff_vp8dsp_init_arm(VP8DSPContext *dsp)
{
    if (HAVE_NEON) {
        dsp->vp8_luma_dc_wht    = ff_vp8_luma_dc_wht_neon;
        dsp->vp8_luma_dc_wht_dc = ff_vp8_luma_dc_wht_dc_neon;

        dsp->vp8_idct_add       = ff_vp8_idct_add_neon;
        dsp->vp8_idct_dc_add    = ff_vp8_idct_dc_add_neon;
        dsp->vp8_idct_dc_add4y  = ff_vp8_idct_dc_add4y_neon;
        dsp->vp8_idct_dc_add4uv = ff_vp8_idct_dc_add4uv_neon;

        dsp->vp8_v_loop_filter16y = ff_vp8_v_loop_filter16_neon;
        dsp->vp8_h_loop_filter16y = ff_vp8_h_loop_filter16_neon;
        dsp->vp8_v_loop_filter8uv = ff_vp8_v_loop_filter8uv_neon;
        dsp->vp8_h_loop_filter8uv = ff_vp8_h_loop_filter8uv_neon;

        dsp->vp8_v_loop_filter16y_inner = ff_vp8_v_loop_filter16_inner_neon;
        dsp->vp8_h_loop_filter16y_inner = ff_vp8_h_loop_filter16_inner_neon;
        dsp->vp8_v_loop_filter8uv_inner = ff_vp8_v_loop_filter8uv_inner_neon;
        dsp->vp8_h_loop_filter8uv_inner = ff_vp8_h_loop_filter8uv_inner_neon;

        dsp->vp8_v_loop_filter_simple = ff_vp8_v_loop_filter16_simple_neon;
        dsp->vp8_h_loop_filter_simple = ff_vp8_h_loop_filter16_simple_neon;

        dsp->put_vp8_epel_pixels_tab[0][0][0] = ff_put_vp8_pixels16_neon;
        dsp->put_vp8_epel_pixels_tab[0][0][2] = ff_put_vp8_epel16_h6_neon;
        dsp->put_vp8_epel_pixels_tab[0][2][0] = ff_put_vp8_epel16_v6_neon;
        dsp->put_vp8_epel_pixels_tab[0][2][2] = ff_put_vp8_epel16_h6v6_neon;

        dsp->put_vp8_epel_pixels_tab[1][0][0] = ff_put_vp8_pixels8_neon;
        dsp->put_vp8_epel_pixels_tab[1][0][1] = ff_put_vp8_epel8_h4_neon;
        dsp->put_vp8_epel_pixels_tab[1][0][2] = ff_put_vp8_epel8_h6_neon;
        dsp->put_vp8_epel_pixels_tab[1][1][0] = ff_put_vp8_epel8_v4_neon;
        dsp->put_vp8_epel_pixels_tab[1][1][1] = ff_put_vp8_epel8_h4v4_neon;
        dsp->put_vp8_epel_pixels_tab[1][1][2] = ff_put_vp8_epel8_h6v4_neon;
        dsp->put_vp8_epel_pixels_tab[1][2][0] = ff_put_vp8_epel8_v6_neon;
        dsp->put_vp8_epel_pixels_tab[1][2][1] = ff_put_vp8_epel8_h4v6_neon;
        dsp->put_vp8_epel_pixels_tab[1][2][2] = ff_put_vp8_epel8_h6v6_neon;

        dsp->put_vp8_epel_pixels_tab[2][0][0] = ff_put_vp8_pixels4_neon;
        dsp->put_vp8_epel_pixels_tab[2][0][1] = ff_put_vp8_epel4_h4_neon;
        dsp->put_vp8_epel_pixels_tab[2][0][2] = ff_put_vp8_epel4_h6_neon;
        dsp->put_vp8_epel_pixels_tab[2][1][0] = ff_put_vp8_epel4_v4_neon;
        dsp->put_vp8_epel_pixels_tab[2][1][1] = ff_put_vp8_epel4_h4v4_neon;
        dsp->put_vp8_epel_pixels_tab[2][1][2] = ff_put_vp8_epel4_h6v4_neon;
        dsp->put_vp8_epel_pixels_tab[2][2][0] = ff_put_vp8_epel4_v6_neon;
        dsp->put_vp8_epel_pixels_tab[2][2][1] = ff_put_vp8_epel4_h4v6_neon;
        dsp->put_vp8_epel_pixels_tab[2][2][2] = ff_put_vp8_epel4_h6v6_neon;

        dsp->put_vp8_bilinear_pixels_tab[0][0][0] = ff_put_vp8_pixels16_neon;
        dsp->put_vp8_bilinear_pixels_tab[0][0][1] = ff_put_vp8_bilin16_h_neon;
        dsp->put_vp8_bilinear_pixels_tab[0][0][2] = ff_put_vp8_bilin16_h_neon;
        dsp->put_vp8_bilinear_pixels_tab[0][1][0] = ff_put_vp8_bilin16_v_neon;
        dsp->put_vp8_bilinear_pixels_tab[0][1][1] = ff_put_vp8_bilin16_hv_neon;
        dsp->put_vp8_bilinear_pixels_tab[0][1][2] = ff_put_vp8_bilin16_hv_neon;
        dsp->put_vp8_bilinear_pixels_tab[0][2][0] = ff_put_vp8_bilin16_v_neon;
        dsp->put_vp8_bilinear_pixels_tab[0][2][1] = ff_put_vp8_bilin16_hv_neon;
        dsp->put_vp8_bilinear_pixels_tab[0][2][2] = ff_put_vp8_bilin16_hv_neon;

        dsp->put_vp8_bilinear_pixels_tab[1][0][0] = ff_put_vp8_pixels8_neon;
        dsp->put_vp8_bilinear_pixels_tab[1][0][1] = ff_put_vp8_bilin8_h_neon;
        dsp->put_vp8_bilinear_pixels_tab[1][0][2] = ff_put_vp8_bilin8_h_neon;
        dsp->put_vp8_bilinear_pixels_tab[1][1][0] = ff_put_vp8_bilin8_v_neon;
        dsp->put_vp8_bilinear_pixels_tab[1][1][1] = ff_put_vp8_bilin8_hv_neon;
        dsp->put_vp8_bilinear_pixels_tab[1][1][2] = ff_put_vp8_bilin8_hv_neon;
        dsp->put_vp8_bilinear_pixels_tab[1][2][0] = ff_put_vp8_bilin8_v_neon;
        dsp->put_vp8_bilinear_pixels_tab[1][2][1] = ff_put_vp8_bilin8_hv_neon;
        dsp->put_vp8_bilinear_pixels_tab[1][2][2] = ff_put_vp8_bilin8_hv_neon;

        dsp->put_vp8_bilinear_pixels_tab[2][0][0] = ff_put_vp8_pixels4_neon;
        dsp->put_vp8_bilinear_pixels_tab[2][0][1] = ff_put_vp8_bilin4_h_neon;
        dsp->put_vp8_bilinear_pixels_tab[2][0][2] = ff_put_vp8_bilin4_h_neon;
        dsp->put_vp8_bilinear_pixels_tab[2][1][0] = ff_put_vp8_bilin4_v_neon;
        dsp->put_vp8_bilinear_pixels_tab[2][1][1] = ff_put_vp8_bilin4_hv_neon;
        dsp->put_vp8_bilinear_pixels_tab[2][1][2] = ff_put_vp8_bilin4_hv_neon;
        dsp->put_vp8_bilinear_pixels_tab[2][2][0] = ff_put_vp8_bilin4_v_neon;
        dsp->put_vp8_bilinear_pixels_tab[2][2][1] = ff_put_vp8_bilin4_hv_neon;
        dsp->put_vp8_bilinear_pixels_tab[2][2][2] = ff_put_vp8_bilin4_hv_neon;
    }
}