summaryrefslogtreecommitdiff
path: root/src/gallium/auxiliary/vl/vp8/reconinter.c
blob: 850a4a74a9c587a49ffba1fe27235e4e8325a440 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
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
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
/*
 *  Copyright (c) 2010 The WebM project authors. All Rights Reserved.
 *
 *  Use of this source code is governed by a BSD-style license
 *  that can be found in the LICENSE file in the root of the source
 *  tree. An additional intellectual property rights grant can be found
 *  in the file PATENTS.  All contributing project authors may
 *  be found in the AUTHORS file in the root of the source tree.
 */


#include "recon_dispatch.h"
#include "filter_dispatch.h"
#include "blockd.h"
#include "reconinter.h"

/* Use this define on systems where unaligned int reads and writes are
 * not allowed, i.e. ARM architectures.
 */
/*#define MUST_BE_ALIGNED*/

static const int bbb[4] = {0, 2, 8, 10};

void vp8_copy_mem16x16_c(unsigned char *src, int src_stride,
                         unsigned char *dst, int dst_stride)
{
    int r;

    for (r = 0; r < 16; r++)
    {
#ifdef MUST_BE_ALIGNED
        dst[0] = src[0];
        dst[1] = src[1];
        dst[2] = src[2];
        dst[3] = src[3];
        dst[4] = src[4];
        dst[5] = src[5];
        dst[6] = src[6];
        dst[7] = src[7];
        dst[8] = src[8];
        dst[9] = src[9];
        dst[10] = src[10];
        dst[11] = src[11];
        dst[12] = src[12];
        dst[13] = src[13];
        dst[14] = src[14];
        dst[15] = src[15];
#else
        ((int *)dst)[0] = ((int *)src)[0] ;
        ((int *)dst)[1] = ((int *)src)[1] ;
        ((int *)dst)[2] = ((int *)src)[2] ;
        ((int *)dst)[3] = ((int *)src)[3] ;
#endif /* MUST_BE_ALIGNED */
        src += src_stride;
        dst += dst_stride;
    }
}

void vp8_copy_mem8x8_c(unsigned char *src, int src_stride,
                       unsigned char *dst, int dst_stride)
{
    int r;

    for (r = 0; r < 8; r++)
    {
#ifdef MUST_BE_ALIGNED
        dst[0] = src[0];
        dst[1] = src[1];
        dst[2] = src[2];
        dst[3] = src[3];
        dst[4] = src[4];
        dst[5] = src[5];
        dst[6] = src[6];
        dst[7] = src[7];
#else
        ((int *)dst)[0] = ((int *)src)[0] ;
        ((int *)dst)[1] = ((int *)src)[1] ;
#endif /* MUST_BE_ALIGNED */
        src += src_stride;
        dst += dst_stride;
    }
}

void vp8_copy_mem8x4_c(unsigned char *src, int src_stride,
                       unsigned char *dst, int dst_stride)
{
    int r;

    for (r = 0; r < 4; r++)
    {
#ifdef MUST_BE_ALIGNED
        dst[0] = src[0];
        dst[1] = src[1];
        dst[2] = src[2];
        dst[3] = src[3];
        dst[4] = src[4];
        dst[5] = src[5];
        dst[6] = src[6];
        dst[7] = src[7];
#else
        ((int *)dst)[0] = ((int *)src)[0] ;
        ((int *)dst)[1] = ((int *)src)[1] ;
#endif /* MUST_BE_ALIGNED */
        src += src_stride;
        dst += dst_stride;
    }
}

void vp8_build_inter_predictors_b(BLOCKD *d, int pitch, vp8_filter_fn_t sppf)
{
    int r;
    unsigned char *ptr_base;
    unsigned char *ptr;
    unsigned char *pred_ptr = d->predictor;

    ptr_base = *(d->base_pre);

    if (d->bmi.mv.as_mv.row & 7 || d->bmi.mv.as_mv.col & 7)
    {
        ptr = ptr_base + d->pre + (d->bmi.mv.as_mv.row >> 3) * d->pre_stride + (d->bmi.mv.as_mv.col >> 3);
        sppf(ptr, d->pre_stride, d->bmi.mv.as_mv.col & 7, d->bmi.mv.as_mv.row & 7, pred_ptr, pitch);
    }
    else
    {
        ptr_base += d->pre + (d->bmi.mv.as_mv.row >> 3) * d->pre_stride + (d->bmi.mv.as_mv.col >> 3);
        ptr = ptr_base;

        for (r = 0; r < 4; r++)
        {
#ifdef MUST_BE_ALIGNED
            pred_ptr[0]  = ptr[0];
            pred_ptr[1]  = ptr[1];
            pred_ptr[2]  = ptr[2];
            pred_ptr[3]  = ptr[3];
#else
            *(int *)pred_ptr = *(int *)ptr ;
#endif /* MUST_BE_ALIGNED */
            pred_ptr     += pitch;
            ptr          += d->pre_stride;
        }
    }
}

static void build_inter_predictors4b(MACROBLOCKD *x, BLOCKD *d, int pitch)
{
    unsigned char *ptr_base;
    unsigned char *ptr;
    unsigned char *pred_ptr = d->predictor;

    ptr_base = *(d->base_pre);
    ptr = ptr_base + d->pre + (d->bmi.mv.as_mv.row >> 3) * d->pre_stride + (d->bmi.mv.as_mv.col >> 3);

    if (d->bmi.mv.as_mv.row & 7 || d->bmi.mv.as_mv.col & 7)
    {
        x->filter_predict8x8(ptr, d->pre_stride, d->bmi.mv.as_mv.col & 7, d->bmi.mv.as_mv.row & 7, pred_ptr, pitch);
    }
    else
    {
        RECON_INVOKE(&x->rtcd->recon, copy8x8)(ptr, d->pre_stride, pred_ptr, pitch);
    }
}

static void build_inter_predictors2b(MACROBLOCKD *x, BLOCKD *d, int pitch)
{
    unsigned char *ptr_base;
    unsigned char *ptr;
    unsigned char *pred_ptr = d->predictor;

    ptr_base = *(d->base_pre);
    ptr = ptr_base + d->pre + (d->bmi.mv.as_mv.row >> 3) * d->pre_stride + (d->bmi.mv.as_mv.col >> 3);

    if (d->bmi.mv.as_mv.row & 7 || d->bmi.mv.as_mv.col & 7)
    {
        x->filter_predict8x4(ptr, d->pre_stride, d->bmi.mv.as_mv.col & 7, d->bmi.mv.as_mv.row & 7, pred_ptr, pitch);
    }
    else
    {
        RECON_INVOKE(&x->rtcd->recon, copy8x4)(ptr, d->pre_stride, pred_ptr, pitch);
    }
}

void vp8_build_inter16x16_predictors_mb(MACROBLOCKD *x,
                                        unsigned char *dst_y,
                                        unsigned char *dst_u,
                                        unsigned char *dst_v,
                                        int dst_ystride,
                                        int dst_uvstride)
{
    int offset;
    unsigned char *ptr;
    unsigned char *uptr, *vptr;

    int mv_row = x->mode_info_context->mbmi.mv.as_mv.row;
    int mv_col = x->mode_info_context->mbmi.mv.as_mv.col;

    unsigned char *ptr_base = x->pre.y_buffer;
    int pre_stride = x->block[0].pre_stride;

    ptr = ptr_base + (mv_row >> 3) * pre_stride + (mv_col >> 3);

    if ((mv_row | mv_col) & 7)
    {
        x->filter_predict16x16(ptr, pre_stride, mv_col & 7, mv_row & 7, dst_y, dst_ystride);
    }
    else
    {
        RECON_INVOKE(&x->rtcd->recon, copy16x16)(ptr, pre_stride, dst_y, dst_ystride);
    }

    mv_row = x->block[16].bmi.mv.as_mv.row;
    mv_col = x->block[16].bmi.mv.as_mv.col;
    pre_stride >>= 1;
    offset = (mv_row >> 3) * pre_stride + (mv_col >> 3);
    uptr = x->pre.u_buffer + offset;
    vptr = x->pre.v_buffer + offset;

    if ((mv_row | mv_col) & 7)
    {
        x->filter_predict8x8(uptr, pre_stride, mv_col & 7, mv_row & 7, dst_u, dst_uvstride);
        x->filter_predict8x8(vptr, pre_stride, mv_col & 7, mv_row & 7, dst_v, dst_uvstride);
    }
    else
    {
        RECON_INVOKE(&x->rtcd->recon, copy8x8)(uptr, pre_stride, dst_u, dst_uvstride);
        RECON_INVOKE(&x->rtcd->recon, copy8x8)(vptr, pre_stride, dst_v, dst_uvstride);
    }
}

void vp8_build_inter4x4_predictors_mb(MACROBLOCKD *x)
{
    int i;

    if (x->mode_info_context->mbmi.partitioning < 3)
    {
        for (i = 0; i < 4; i++)
        {
            BLOCKD *d = &x->block[bbb[i]];
            build_inter_predictors4b(x, d, 16);
        }
    }
    else
    {
        for (i = 0; i < 16; i += 2)
        {
            BLOCKD *d0 = &x->block[i];
            BLOCKD *d1 = &x->block[i+1];

            if (d0->bmi.mv.as_int == d1->bmi.mv.as_int)
                build_inter_predictors2b(x, d0, 16);
            else
            {
                vp8_build_inter_predictors_b(d0, 16, x->filter_predict4x4);
                vp8_build_inter_predictors_b(d1, 16, x->filter_predict4x4);
            }
        }
    }

    for (i = 16; i < 24; i += 2)
    {
        BLOCKD *d0 = &x->block[i];
        BLOCKD *d1 = &x->block[i+1];

        if (d0->bmi.mv.as_int == d1->bmi.mv.as_int)
            build_inter_predictors2b(x, d0, 8);
        else
        {
            vp8_build_inter_predictors_b(d0, 8, x->filter_predict4x4);
            vp8_build_inter_predictors_b(d1, 8, x->filter_predict4x4);
        }
    }
}

void vp8_build_inter_predictors_mb(MACROBLOCKD *x)
{
    if (x->mode_info_context->mbmi.mode != SPLITMV)
    {
        vp8_build_inter16x16_predictors_mb(x, x->predictor, &x->predictor[256],
                                           &x->predictor[320], 16, 8);
    }
    else
    {
        vp8_build_inter4x4_predictors_mb(x);
    }
}

void vp8_build_uvmvs(MACROBLOCKD *x, int fullpixel)
{
    int i, j;

    if (x->mode_info_context->mbmi.mode == SPLITMV)
    {
        for (i = 0; i < 2; i++)
        {
            for (j = 0; j < 2; j++)
            {
                int yoffset = i * 8 + j * 2;
                int uoffset = 16 + i * 2 + j;
                int voffset = 20 + i * 2 + j;

                int temp;

                temp = x->block[yoffset  ].bmi.mv.as_mv.row
                       + x->block[yoffset+1].bmi.mv.as_mv.row
                       + x->block[yoffset+4].bmi.mv.as_mv.row
                       + x->block[yoffset+5].bmi.mv.as_mv.row;

                if (temp < 0) temp -= 4;
                else temp += 4;

                x->block[uoffset].bmi.mv.as_mv.row = temp / 8;

                if (fullpixel)
                    x->block[uoffset].bmi.mv.as_mv.row = (temp / 8) & 0xfffffff8;

                temp = x->block[yoffset  ].bmi.mv.as_mv.col
                       + x->block[yoffset+1].bmi.mv.as_mv.col
                       + x->block[yoffset+4].bmi.mv.as_mv.col
                       + x->block[yoffset+5].bmi.mv.as_mv.col;

                if (temp < 0) temp -= 4;
                else temp += 4;

                x->block[uoffset].bmi.mv.as_mv.col = temp / 8;

                if (fullpixel)
                    x->block[uoffset].bmi.mv.as_mv.col = (temp / 8) & 0xfffffff8;

                x->block[voffset].bmi.mv.as_mv.row = x->block[uoffset].bmi.mv.as_mv.row ;
                x->block[voffset].bmi.mv.as_mv.col = x->block[uoffset].bmi.mv.as_mv.col ;
            }
        }
    }
    else
    {
        int mvrow = x->mode_info_context->mbmi.mv.as_mv.row;
        int mvcol = x->mode_info_context->mbmi.mv.as_mv.col;

        if (mvrow < 0)
            mvrow -= 1;
        else
            mvrow += 1;

        if (mvcol < 0)
            mvcol -= 1;
        else
            mvcol += 1;

        mvrow /= 2;
        mvcol /= 2;

        for (i = 0; i < 8; i++)
        {
            x->block[ 16 + i].bmi.mv.as_mv.row = mvrow;
            x->block[ 16 + i].bmi.mv.as_mv.col = mvcol;

            if (fullpixel)
            {
                x->block[ 16 + i].bmi.mv.as_mv.row = mvrow & 0xfffffff8;
                x->block[ 16 + i].bmi.mv.as_mv.col = mvcol & 0xfffffff8;
            }
        }
    }
}