summaryrefslogtreecommitdiff
path: root/libavcodec/arm/h264pred_neon.S
blob: 63c96ee725258b9b33d955342f4786f104ffe22f (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
/*
 * Copyright (c) 2009 Mans Rullgard <mans@mansr.com>
 *
 * 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 "asm.S"

        .macro ldcol.8  rd,  rs,  rt,  n=8,  hi=0
.if \n == 8 || \hi == 0
        vld1.8          {\rd[0]}, [\rs], \rt
        vld1.8          {\rd[1]}, [\rs], \rt
        vld1.8          {\rd[2]}, [\rs], \rt
        vld1.8          {\rd[3]}, [\rs], \rt
.endif
.if \n == 8 || \hi == 1
        vld1.8          {\rd[4]}, [\rs], \rt
        vld1.8          {\rd[5]}, [\rs], \rt
        vld1.8          {\rd[6]}, [\rs], \rt
        vld1.8          {\rd[7]}, [\rs], \rt
.endif
        .endm

        .macro add16x8  dq,  dl,  dh,  rl,  rh
        vaddl.u8        \dq, \rl, \rh
        vadd.u16        \dl, \dl, \dh
        vpadd.u16       \dl, \dl, \dl
        vpadd.u16       \dl, \dl, \dl
        .endm

function ff_pred16x16_128_dc_neon, export=1
        vmov.i8         q0,  #128
        b               .L_pred16x16_dc_end
endfunc

function ff_pred16x16_top_dc_neon, export=1
        sub             r2,  r0,  r1
        vld1.8          {q0},     [r2,:128]
        add16x8         q0,  d0,  d1,  d0,  d1
        vrshrn.u16      d0,  q0,  #4
        vdup.8          q0,  d0[0]
        b               .L_pred16x16_dc_end
endfunc

function ff_pred16x16_left_dc_neon, export=1
        sub             r2,  r0,  #1
        ldcol.8         d0,  r2,  r1
        ldcol.8         d1,  r2,  r1
        add16x8         q0,  d0,  d1,  d0,  d1
        vrshrn.u16      d0,  q0,  #4
        vdup.8          q0,  d0[0]
        b               .L_pred16x16_dc_end
endfunc

function ff_pred16x16_dc_neon, export=1
        sub             r2,  r0,  r1
        vld1.8          {q0},     [r2,:128]
        sub             r2,  r0,  #1
        ldcol.8         d2,  r2,  r1
        ldcol.8         d3,  r2,  r1
        vaddl.u8        q0,  d0,  d1
        vaddl.u8        q1,  d2,  d3
        vadd.u16        q0,  q0,  q1
        vadd.u16        d0,  d0,  d1
        vpadd.u16       d0,  d0,  d0
        vpadd.u16       d0,  d0,  d0
        vrshrn.u16      d0,  q0,  #5
        vdup.8          q0,  d0[0]
.L_pred16x16_dc_end:
        mov             r3,  #8
6:      vst1.8          {q0},     [r0,:128], r1
        vst1.8          {q0},     [r0,:128], r1
        subs            r3,  r3,  #1
        bne             6b
        bx              lr
endfunc

function ff_pred16x16_hor_neon, export=1
        sub             r2,  r0,  #1
        mov             r3,  #16
1:      vld1.8          {d0[],d1[]},[r2],      r1
        vst1.8          {q0},       [r0,:128], r1
        subs            r3,  r3,  #1
        bne             1b
        bx              lr
endfunc

function ff_pred16x16_vert_neon, export=1
        sub             r0,  r0,  r1
        vld1.8          {q0},     [r0,:128], r1
        mov             r3,  #8
1:      vst1.8          {q0},     [r0,:128], r1
        vst1.8          {q0},     [r0,:128], r1
        subs            r3,  r3,  #1
        bne             1b
        bx              lr
endfunc

function ff_pred16x16_plane_neon, export=1
        sub             r3,  r0,  r1
        add             r2,  r3,  #8
        sub             r3,  r3,  #1
        vld1.8          {d0},     [r3]
        vld1.8          {d2},     [r2,:64], r1
        ldcol.8         d1,  r3,  r1
        add             r3,  r3,  r1
        ldcol.8         d3,  r3,  r1
        vrev64.8        q0,  q0
        vaddl.u8        q8,  d2,  d3
        vsubl.u8        q2,  d2,  d0
        vsubl.u8        q3,  d3,  d1
        movrel          r3,  p16weight
        vld1.8          {q0},     [r3,:128]
        vmul.s16        q2,  q2,  q0
        vmul.s16        q3,  q3,  q0
        vadd.i16        d4,  d4,  d5
        vadd.i16        d5,  d6,  d7
        vpadd.i16       d4,  d4,  d5
        vpadd.i16       d4,  d4,  d4
        vshll.s16       q3,  d4,  #2
        vaddw.s16       q2,  q3,  d4
        vrshrn.s32      d4,  q2,  #6
        mov             r3,  #0
        vtrn.16         d4,  d5
        vadd.i16        d2,  d4,  d5
        vshl.i16        d3,  d2,  #3
        vrev64.16       d16, d17
        vsub.i16        d3,  d3,  d2
        vadd.i16        d16, d16, d0
        vshl.i16        d2,  d16, #4
        vsub.i16        d2,  d2,  d3
        vshl.i16        d3,  d4,  #4
        vext.16         q0,  q0,  q0,  #7
        vsub.i16        d6,  d5,  d3
        vmov.16         d0[0], r3
        vmul.i16        q0,  q0,  d4[0]
        vdup.16         q1,  d2[0]
        vdup.16         q2,  d4[0]
        vdup.16         q3,  d6[0]
        vshl.i16        q2,  q2,  #3
        vadd.i16        q1,  q1,  q0
        vadd.i16        q3,  q3,  q2
        mov             r3,  #16
1:
        vqshrun.s16     d0,  q1,  #5
        vadd.i16        q1,  q1,  q2
        vqshrun.s16     d1,  q1,  #5
        vadd.i16        q1,  q1,  q3
        vst1.8          {q0},     [r0,:128], r1
        subs            r3,  r3,  #1
        bne             1b
        bx              lr
endfunc

        .section        .rodata
        .align          4
p16weight:
        .short          1,2,3,4,5,6,7,8

        .text

function ff_pred8x8_hor_neon, export=1
        sub             r2,  r0,  #1
        mov             r3,  #8
1:      vld1.8          {d0[]},   [r2],     r1
        vst1.8          {d0},     [r0,:64], r1
        subs            r3,  r3,  #1
        bne             1b
        bx              lr
endfunc

function ff_pred8x8_vert_neon, export=1
        sub             r0,  r0,  r1
        vld1.8          {d0},     [r0,:64], r1
        mov             r3,  #4
1:      vst1.8          {d0},     [r0,:64], r1
        vst1.8          {d0},     [r0,:64], r1
        subs            r3,  r3,  #1
        bne             1b
        bx              lr
endfunc

function ff_pred8x8_plane_neon, export=1
        sub             r3,  r0,  r1
        add             r2,  r3,  #4
        sub             r3,  r3,  #1
        vld1.32         {d0[0]},  [r3]
        vld1.32         {d2[0]},  [r2,:32], r1
        ldcol.8         d0,  r3,  r1,  4,  hi=1
        add             r3,  r3,  r1
        ldcol.8         d3,  r3,  r1,  4
        vaddl.u8        q8,  d2,  d3
        vrev32.8        d0,  d0
        vtrn.32         d2,  d3
        vsubl.u8        q2,  d2,  d0
        movrel          r3,  p16weight
        vld1.16         {q0},     [r3,:128]
        vmul.s16        d4,  d4,  d0
        vmul.s16        d5,  d5,  d0
        vpadd.i16       d4,  d4,  d5
        vpaddl.s16      d4,  d4
        vshl.i32        d5,  d4,  #4
        vadd.s32        d4,  d4,  d5
        vrshrn.s32      d4,  q2,  #5
        mov             r3,  #0
        vtrn.16         d4,  d5
        vadd.i16        d2,  d4,  d5
        vshl.i16        d3,  d2,  #2
        vrev64.16       d16, d16
        vsub.i16        d3,  d3,  d2
        vadd.i16        d16, d16, d0
        vshl.i16        d2,  d16, #4
        vsub.i16        d2,  d2,  d3
        vshl.i16        d3,  d4,  #3
        vext.16         q0,  q0,  q0,  #7
        vsub.i16        d6,  d5,  d3
        vmov.16         d0[0], r3
        vmul.i16        q0,  q0,  d4[0]
        vdup.16         q1,  d2[0]
        vdup.16         q2,  d4[0]
        vdup.16         q3,  d6[0]
        vshl.i16        q2,  q2,  #3
        vadd.i16        q1,  q1,  q0
        vadd.i16        q3,  q3,  q2
        mov             r3,  #8
1:
        vqshrun.s16     d0,  q1,  #5
        vadd.i16        q1,  q1,  q3
        vst1.8          {d0},     [r0,:64], r1
        subs            r3,  r3,  #1
        bne             1b
        bx              lr
endfunc

function ff_pred8x8_128_dc_neon, export=1
        vmov.i8         q0,  #128
        b               .L_pred8x8_dc_end
endfunc

function ff_pred8x8_top_dc_neon, export=1
        sub             r2,  r0,  r1
        vld1.8          {d0},     [r2,:64]
        vpaddl.u8       d0,  d0
        vpadd.u16       d0,  d0,  d0
        vrshrn.u16      d0,  q0,  #2
        vdup.8          d1,  d0[1]
        vdup.8          d0,  d0[0]
        vtrn.32         d0,  d1
        b               .L_pred8x8_dc_end
endfunc

function ff_pred8x8_left_dc_neon, export=1
        sub             r2,  r0,  #1
        ldcol.8         d0,  r2,  r1
        vpaddl.u8       d0,  d0
        vpadd.u16       d0,  d0,  d0
        vrshrn.u16      d0,  q0,  #2
        vdup.8          d1,  d0[1]
        vdup.8          d0,  d0[0]
        b               .L_pred8x8_dc_end
endfunc

function ff_pred8x8_dc_neon, export=1
        sub             r2,  r0,  r1
        vld1.8          {d0},     [r2,:64]
        sub             r2,  r0,  #1
        ldcol.8         d1,  r2,  r1
        vtrn.32         d0,  d1
        vpaddl.u8       q0,  q0
        vpadd.u16       d0,  d0,  d1
        vpadd.u16       d1,  d0,  d0
        vrshrn.u16      d2,  q0,  #3
        vrshrn.u16      d3,  q0,  #2
        vdup.8          d0,  d2[4]
        vdup.8          d1,  d3[3]
        vdup.8          d4,  d3[2]
        vdup.8          d5,  d2[5]
        vtrn.32         q0,  q2
.L_pred8x8_dc_end:
        mov             r3,  #4
        add             r2,  r0,  r1,  lsl #2
6:      vst1.8          {d0},     [r0,:64], r1
        vst1.8          {d1},     [r2,:64], r1
        subs            r3,  r3,  #1
        bne             6b
        bx              lr
endfunc

function ff_pred8x8_l0t_dc_neon, export=1
        sub             r2,  r0,  r1
        vld1.8          {d0},     [r2,:64]
        sub             r2,  r0,  #1
        ldcol.8         d1,  r2,  r1,  4
        vtrn.32         d0,  d1
        vpaddl.u8       q0,  q0
        vpadd.u16       d0,  d0,  d1
        vpadd.u16       d1,  d0,  d0
        vrshrn.u16      d2,  q0,  #3
        vrshrn.u16      d3,  q0,  #2
        vdup.8          d0,  d2[4]
        vdup.8          d1,  d3[0]
        vdup.8          q2,  d3[2]
        vtrn.32         q0,  q2
        b               .L_pred8x8_dc_end
endfunc

function ff_pred8x8_l00_dc_neon, export=1
        sub             r2,  r0,  #1
        ldcol.8         d0,  r2,  r1,  4
        vpaddl.u8       d0,  d0
        vpadd.u16       d0,  d0,  d0
        vrshrn.u16      d0,  q0,  #2
        vmov.i8         d1,  #128
        vdup.8          d0,  d0[0]
        b               .L_pred8x8_dc_end
endfunc

function ff_pred8x8_0lt_dc_neon, export=1
        sub             r2,  r0,  r1
        vld1.8          {d0},     [r2,:64]
        add             r2,  r0,  r1,  lsl #2
        sub             r2,  r2,  #1
        ldcol.8         d1,  r2,  r1,  4,  hi=1
        vtrn.32         d0,  d1
        vpaddl.u8       q0,  q0
        vpadd.u16       d0,  d0,  d1
        vpadd.u16       d1,  d0,  d0
        vrshrn.u16      d3,  q0,  #2
        vrshrn.u16      d2,  q0,  #3
        vdup.8          d0,  d3[0]
        vdup.8          d1,  d3[3]
        vdup.8          d4,  d3[2]
        vdup.8          d5,  d2[5]
        vtrn.32         q0,  q2
        b               .L_pred8x8_dc_end
endfunc

function ff_pred8x8_0l0_dc_neon, export=1
        add             r2,  r0,  r1,  lsl #2
        sub             r2,  r2,  #1
        ldcol.8         d1,  r2,  r1,  4
        vpaddl.u8       d2,  d1
        vpadd.u16       d2,  d2,  d2
        vrshrn.u16      d1,  q1,  #2
        vmov.i8         d0,  #128
        vdup.8          d1,  d1[0]
        b               .L_pred8x8_dc_end
endfunc