[deb_ffmpeg.git] / ffmpeg / libavcodec / h264_mc_template.c

/*
 * H.26L/H.264/AVC/JVT/14496-10/... decoder
 * Copyright (c) 2003 Michael Niedermayer <michaelni@gmx.at>
 *
 * This file is part of FFmpeg.
 *
 * FFmpeg is free software; you can redistribute it and/or
 * 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.
 *
 * FFmpeg is distributed in the hope that it will be useful,
 * 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
 * License along with FFmpeg; if not, write to the Free Software
 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
 */

#include "h264.h"

#undef MCFUNC

#if   CHROMA_IDC == 1
#   define MCFUNC(n) FUNC(n ## _420)
#elif CHROMA_IDC == 2
#   define MCFUNC(n) FUNC(n ## _422)
#elif CHROMA_IDC == 3
#   define MCFUNC(n) FUNC(n ## _444)
#endif

#undef  mc_part
#define mc_part MCFUNC(mc_part)

static void mc_part(H264Context *h, int n, int square,
                    int height, int delta,
                    uint8_t *dest_y, uint8_t *dest_cb,
                    uint8_t *dest_cr,
                    int x_offset, int y_offset,
                    qpel_mc_func *qpix_put,
                    h264_chroma_mc_func chroma_put,
                    qpel_mc_func *qpix_avg,
                    h264_chroma_mc_func chroma_avg,
                    h264_weight_func *weight_op,
                    h264_biweight_func *weight_avg,
                    int list0, int list1)
{
    if ((h->use_weight == 2 && list0 && list1 &&
         (h->implicit_weight[h->ref_cache[0][scan8[n]]][h->ref_cache[1][scan8[n]]][h->mb_y & 1] != 32)) ||
        h->use_weight == 1)
        mc_part_weighted(h, n, square, height, delta, dest_y, dest_cb, dest_cr,
                         x_offset, y_offset, qpix_put, chroma_put,
                         weight_op[0], weight_op[1], weight_avg[0],
                         weight_avg[1], list0, list1, PIXEL_SHIFT, CHROMA_IDC);
    else
        mc_part_std(h, n, square, height, delta, dest_y, dest_cb, dest_cr,
                    x_offset, y_offset, qpix_put, chroma_put, qpix_avg,
                    chroma_avg, list0, list1, PIXEL_SHIFT, CHROMA_IDC);
}

static void MCFUNC(hl_motion)(H264Context *h, uint8_t *dest_y,
                              uint8_t *dest_cb, uint8_t *dest_cr,
                              qpel_mc_func(*qpix_put)[16],
                              h264_chroma_mc_func(*chroma_put),
                              qpel_mc_func(*qpix_avg)[16],
                              h264_chroma_mc_func(*chroma_avg),
                              h264_weight_func *weight_op,
                              h264_biweight_func *weight_avg)
{
    const int mb_xy   = h->mb_xy;
    const int mb_type = h->cur_pic.mb_type[mb_xy];

    av_assert2(IS_INTER(mb_type));

    if (HAVE_THREADS && (h->avctx->active_thread_type & FF_THREAD_FRAME))
        await_references(h);
    prefetch_motion(h, 0, PIXEL_SHIFT, CHROMA_IDC);

    if (IS_16X16(mb_type)) {
        mc_part(h, 0, 1, 16, 0, dest_y, dest_cb, dest_cr, 0, 0,
                qpix_put[0], chroma_put[0], qpix_avg[0], chroma_avg[0],
                weight_op, weight_avg,
                IS_DIR(mb_type, 0, 0), IS_DIR(mb_type, 0, 1));
    } else if (IS_16X8(mb_type)) {
        mc_part(h, 0, 0, 8, 8 << PIXEL_SHIFT, dest_y, dest_cb, dest_cr, 0, 0,
                qpix_put[1], chroma_put[0], qpix_avg[1], chroma_avg[0],
                weight_op, weight_avg,
                IS_DIR(mb_type, 0, 0), IS_DIR(mb_type, 0, 1));
        mc_part(h, 8, 0, 8, 8 << PIXEL_SHIFT, dest_y, dest_cb, dest_cr, 0, 4,
                qpix_put[1], chroma_put[0], qpix_avg[1], chroma_avg[0],
                weight_op, weight_avg,
                IS_DIR(mb_type, 1, 0), IS_DIR(mb_type, 1, 1));
    } else if (IS_8X16(mb_type)) {
        mc_part(h, 0, 0, 16, 8 * h->mb_linesize, dest_y, dest_cb, dest_cr, 0, 0,
                qpix_put[1], chroma_put[1], qpix_avg[1], chroma_avg[1],
                &weight_op[1], &weight_avg[1],
                IS_DIR(mb_type, 0, 0), IS_DIR(mb_type, 0, 1));
        mc_part(h, 4, 0, 16, 8 * h->mb_linesize, dest_y, dest_cb, dest_cr, 4, 0,
                qpix_put[1], chroma_put[1], qpix_avg[1], chroma_avg[1],
                &weight_op[1], &weight_avg[1],
                IS_DIR(mb_type, 1, 0), IS_DIR(mb_type, 1, 1));
    } else {
        int i;

        av_assert2(IS_8X8(mb_type));

        for (i = 0; i < 4; i++) {
            const int sub_mb_type = h->sub_mb_type[i];
            const int n  = 4 * i;
            int x_offset = (i & 1) << 2;
            int y_offset = (i & 2) << 1;

            if (IS_SUB_8X8(sub_mb_type)) {
                mc_part(h, n, 1, 8, 0, dest_y, dest_cb, dest_cr,
                        x_offset, y_offset,
                        qpix_put[1], chroma_put[1], qpix_avg[1], chroma_avg[1],
                        &weight_op[1], &weight_avg[1],
                        IS_DIR(sub_mb_type, 0, 0), IS_DIR(sub_mb_type, 0, 1));
            } else if (IS_SUB_8X4(sub_mb_type)) {
                mc_part(h, n, 0, 4, 4 << PIXEL_SHIFT, dest_y, dest_cb, dest_cr,
                        x_offset, y_offset,
                        qpix_put[2], chroma_put[1], qpix_avg[2], chroma_avg[1],
                        &weight_op[1], &weight_avg[1],
                        IS_DIR(sub_mb_type, 0, 0), IS_DIR(sub_mb_type, 0, 1));
                mc_part(h, n + 2, 0, 4, 4 << PIXEL_SHIFT,
                        dest_y, dest_cb, dest_cr, x_offset, y_offset + 2,
                        qpix_put[2], chroma_put[1], qpix_avg[2], chroma_avg[1],
                        &weight_op[1], &weight_avg[1],
                        IS_DIR(sub_mb_type, 0, 0), IS_DIR(sub_mb_type, 0, 1));
            } else if (IS_SUB_4X8(sub_mb_type)) {
                mc_part(h, n, 0, 8, 4 * h->mb_linesize,
                        dest_y, dest_cb, dest_cr, x_offset, y_offset,
                        qpix_put[2], chroma_put[2], qpix_avg[2], chroma_avg[2],
                        &weight_op[2], &weight_avg[2],
                        IS_DIR(sub_mb_type, 0, 0), IS_DIR(sub_mb_type, 0, 1));
                mc_part(h, n + 1, 0, 8, 4 * h->mb_linesize,
                        dest_y, dest_cb, dest_cr, x_offset + 2, y_offset,
                        qpix_put[2], chroma_put[2], qpix_avg[2], chroma_avg[2],
                        &weight_op[2], &weight_avg[2],
                        IS_DIR(sub_mb_type, 0, 0), IS_DIR(sub_mb_type, 0, 1));
            } else {
                int j;
                av_assert2(IS_SUB_4X4(sub_mb_type));
                for (j = 0; j < 4; j++) {
                    int sub_x_offset = x_offset + 2 * (j & 1);
                    int sub_y_offset = y_offset + (j & 2);
                    mc_part(h, n + j, 1, 4, 0,
                            dest_y, dest_cb, dest_cr, sub_x_offset, sub_y_offset,
                            qpix_put[2], chroma_put[2], qpix_avg[2], chroma_avg[2],
                            &weight_op[2], &weight_avg[2],
                            IS_DIR(sub_mb_type, 0, 0), IS_DIR(sub_mb_type, 0, 1));
                }
            }
        }
    }

    prefetch_motion(h, 1, PIXEL_SHIFT, CHROMA_IDC);
}
Commit	Line	Data
2ba45a60 DM	1	/*
	2	* H.26L/H.264/AVC/JVT/14496-10/... decoder
	3	* Copyright (c) 2003 Michael Niedermayer <michaelni@gmx.at>
	4	*
	5	* This file is part of FFmpeg.
	6	*
	7	* FFmpeg is free software; you can redistribute it and/or
	8	* modify it under the terms of the GNU Lesser General Public
	9	* License as published by the Free Software Foundation; either
	10	* version 2.1 of the License, or (at your option) any later version.
	11	*
	12	* FFmpeg is distributed in the hope that it will be useful,
	13	* but WITHOUT ANY WARRANTY; without even the implied warranty of
	14	* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
	15	* Lesser General Public License for more details.
	16	*
	17	* You should have received a copy of the GNU Lesser General Public
	18	* License along with FFmpeg; if not, write to the Free Software
	19	* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
	20	*/
	21
	22	#include "h264.h"
	23
	24	#undef MCFUNC
	25
	26	#if CHROMA_IDC == 1
	27	# define MCFUNC(n) FUNC(n ## _420)
	28	#elif CHROMA_IDC == 2
	29	# define MCFUNC(n) FUNC(n ## _422)
	30	#elif CHROMA_IDC == 3
	31	# define MCFUNC(n) FUNC(n ## _444)
	32	#endif
	33
	34	#undef mc_part
	35	#define mc_part MCFUNC(mc_part)
	36
	37	static void mc_part(H264Context *h, int n, int square,
	38	int height, int delta,
	39	uint8_t dest_y, uint8_t dest_cb,
	40	uint8_t *dest_cr,
	41	int x_offset, int y_offset,
	42	qpel_mc_func *qpix_put,
	43	h264_chroma_mc_func chroma_put,
	44	qpel_mc_func *qpix_avg,
	45	h264_chroma_mc_func chroma_avg,
	46	h264_weight_func *weight_op,
	47	h264_biweight_func *weight_avg,
	48	int list0, int list1)
	49	{
	50	if ((h->use_weight == 2 && list0 && list1 &&
	51	(h->implicit_weight[h->ref_cache[0][scan8[n]]][h->ref_cache[1][scan8[n]]][h->mb_y & 1] != 32)) \|\|
	52	h->use_weight == 1)
	53	mc_part_weighted(h, n, square, height, delta, dest_y, dest_cb, dest_cr,
	54	x_offset, y_offset, qpix_put, chroma_put,
	55	weight_op[0], weight_op[1], weight_avg[0],
	56	weight_avg[1], list0, list1, PIXEL_SHIFT, CHROMA_IDC);
	57	else
	58	mc_part_std(h, n, square, height, delta, dest_y, dest_cb, dest_cr,
	59	x_offset, y_offset, qpix_put, chroma_put, qpix_avg,
	60	chroma_avg, list0, list1, PIXEL_SHIFT, CHROMA_IDC);
	61	}
	62
	63	static void MCFUNC(hl_motion)(H264Context h, uint8_t dest_y,
	64	uint8_t dest_cb, uint8_t dest_cr,
65	qpel_mc_func(*qpix_put)[16],
66	h264_chroma_mc_func(*chroma_put),
67	qpel_mc_func(*qpix_avg)[16],
68	h264_chroma_mc_func(*chroma_avg),
69	h264_weight_func *weight_op,
70	h264_biweight_func *weight_avg)
71	{
72	const int mb_xy = h->mb_xy;
73	const int mb_type = h->cur_pic.mb_type[mb_xy];
74
75	av_assert2(IS_INTER(mb_type));
76
77	if (HAVE_THREADS && (h->avctx->active_thread_type & FF_THREAD_FRAME))
78	await_references(h);
79	prefetch_motion(h, 0, PIXEL_SHIFT, CHROMA_IDC);
80
81	if (IS_16X16(mb_type)) {
82	mc_part(h, 0, 1, 16, 0, dest_y, dest_cb, dest_cr, 0, 0,
83	qpix_put[0], chroma_put[0], qpix_avg[0], chroma_avg[0],
84	weight_op, weight_avg,
85	IS_DIR(mb_type, 0, 0), IS_DIR(mb_type, 0, 1));
86	} else if (IS_16X8(mb_type)) {
87	mc_part(h, 0, 0, 8, 8 << PIXEL_SHIFT, dest_y, dest_cb, dest_cr, 0, 0,
88	qpix_put[1], chroma_put[0], qpix_avg[1], chroma_avg[0],
89	weight_op, weight_avg,
90	IS_DIR(mb_type, 0, 0), IS_DIR(mb_type, 0, 1));
91	mc_part(h, 8, 0, 8, 8 << PIXEL_SHIFT, dest_y, dest_cb, dest_cr, 0, 4,
92	qpix_put[1], chroma_put[0], qpix_avg[1], chroma_avg[0],
93	weight_op, weight_avg,
94	IS_DIR(mb_type, 1, 0), IS_DIR(mb_type, 1, 1));
95	} else if (IS_8X16(mb_type)) {
96	mc_part(h, 0, 0, 16, 8 * h->mb_linesize, dest_y, dest_cb, dest_cr, 0, 0,
97	qpix_put[1], chroma_put[1], qpix_avg[1], chroma_avg[1],
98	&weight_op[1], &weight_avg[1],
99	IS_DIR(mb_type, 0, 0), IS_DIR(mb_type, 0, 1));
100	mc_part(h, 4, 0, 16, 8 * h->mb_linesize, dest_y, dest_cb, dest_cr, 4, 0,
101	qpix_put[1], chroma_put[1], qpix_avg[1], chroma_avg[1],
102	&weight_op[1], &weight_avg[1],
103	IS_DIR(mb_type, 1, 0), IS_DIR(mb_type, 1, 1));
104	} else {
105	int i;
106
107	av_assert2(IS_8X8(mb_type));
108
109	for (i = 0; i < 4; i++) {
110	const int sub_mb_type = h->sub_mb_type[i];
111	const int n = 4 * i;
112	int x_offset = (i & 1) << 2;
113	int y_offset = (i & 2) << 1;
114
115	if (IS_SUB_8X8(sub_mb_type)) {
116	mc_part(h, n, 1, 8, 0, dest_y, dest_cb, dest_cr,
117	x_offset, y_offset,
118	qpix_put[1], chroma_put[1], qpix_avg[1], chroma_avg[1],
119	&weight_op[1], &weight_avg[1],
120	IS_DIR(sub_mb_type, 0, 0), IS_DIR(sub_mb_type, 0, 1));
121	} else if (IS_SUB_8X4(sub_mb_type)) {
122	mc_part(h, n, 0, 4, 4 << PIXEL_SHIFT, dest_y, dest_cb, dest_cr,
123	x_offset, y_offset,
124	qpix_put[2], chroma_put[1], qpix_avg[2], chroma_avg[1],
125	&weight_op[1], &weight_avg[1],
126	IS_DIR(sub_mb_type, 0, 0), IS_DIR(sub_mb_type, 0, 1));
127	mc_part(h, n + 2, 0, 4, 4 << PIXEL_SHIFT,
128	dest_y, dest_cb, dest_cr, x_offset, y_offset + 2,
129	qpix_put[2], chroma_put[1], qpix_avg[2], chroma_avg[1],
130	&weight_op[1], &weight_avg[1],
131	IS_DIR(sub_mb_type, 0, 0), IS_DIR(sub_mb_type, 0, 1));
132	} else if (IS_SUB_4X8(sub_mb_type)) {
133	mc_part(h, n, 0, 8, 4 * h->mb_linesize,
134	dest_y, dest_cb, dest_cr, x_offset, y_offset,
135	qpix_put[2], chroma_put[2], qpix_avg[2], chroma_avg[2],
136	&weight_op[2], &weight_avg[2],
137	IS_DIR(sub_mb_type, 0, 0), IS_DIR(sub_mb_type, 0, 1));
138	mc_part(h, n + 1, 0, 8, 4 * h->mb_linesize,
139	dest_y, dest_cb, dest_cr, x_offset + 2, y_offset,
140	qpix_put[2], chroma_put[2], qpix_avg[2], chroma_avg[2],
141	&weight_op[2], &weight_avg[2],
142	IS_DIR(sub_mb_type, 0, 0), IS_DIR(sub_mb_type, 0, 1));
143	} else {
144	int j;
145	av_assert2(IS_SUB_4X4(sub_mb_type));
146	for (j = 0; j < 4; j++) {
147	int sub_x_offset = x_offset + 2 * (j & 1);
148	int sub_y_offset = y_offset + (j & 2);
149	mc_part(h, n + j, 1, 4, 0,
150	dest_y, dest_cb, dest_cr, sub_x_offset, sub_y_offset,
151	qpix_put[2], chroma_put[2], qpix_avg[2], chroma_avg[2],
152	&weight_op[2], &weight_avg[2],
153	IS_DIR(sub_mb_type, 0, 0), IS_DIR(sub_mb_type, 0, 1));
154	}
155	}
156	}
157	}
158
159	prefetch_motion(h, 1, PIXEL_SHIFT, CHROMA_IDC);
160	}
161