aboutsummaryrefslogtreecommitdiffstats
path: root/libavcodec/aarch64/h264qpel_neon.S
blob: d27cfac4941d6ea564f99c0acf19af3cfc8aae72 (plain) (blame)
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
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
/*
 * Copyright (c) 2008 Mans Rullgard <mans@mansr.com>
 * Copyright (c) 2013 Janne Grunau <janne-libav@jannau.net>
 *
 * 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 "libavutil/aarch64/asm.S"
#include "neon.S"

        /* H.264 qpel MC */

.macro  lowpass_const   r
        movz            \r, #20, lsl #16
        movk            \r, #5
        mov             v6.S[0], \r
.endm

//trashes v0-v5
.macro  lowpass_8       r0,  r1,  r2,  r3,  d0,  d1,  narrow=1
        ext             v2.8B,      \r0\().8B, \r1\().8B, #2
        ext             v3.8B,      \r0\().8B, \r1\().8B, #3
        uaddl           v2.8H,      v2.8B,     v3.8B
        ext             v4.8B,      \r0\().8B, \r1\().8B, #1
        ext             v5.8B,      \r0\().8B, \r1\().8B, #4
        uaddl           v4.8H,      v4.8B,     v5.8B
        ext             v1.8B,      \r0\().8B, \r1\().8B, #5
        uaddl           \d0\().8H,  \r0\().8B, v1.8B
        ext             v0.8B,      \r2\().8B, \r3\().8B, #2
        mla             \d0\().8H,  v2.8H,     v6.H[1]
        ext             v1.8B,      \r2\().8B, \r3\().8B, #3
        uaddl           v0.8H,      v0.8B,     v1.8B
        ext             v1.8B,      \r2\().8B, \r3\().8B, #1
        mls             \d0\().8H,  v4.8H,     v6.H[0]
        ext             v3.8B,      \r2\().8B, \r3\().8B, #4
        uaddl           v1.8H,      v1.8B,     v3.8B
        ext             v2.8B,      \r2\().8B, \r3\().8B, #5
        uaddl           \d1\().8H,  \r2\().8B, v2.8B
        mla             \d1\().8H,  v0.8H,     v6.H[1]
        mls             \d1\().8H,  v1.8H,     v6.H[0]
  .if \narrow
        sqrshrun        \d0\().8B,  \d0\().8H, #5
        sqrshrun        \d1\().8B,  \d1\().8H, #5
  .endif
.endm

//trashes v0-v5, v7, v30-v31
.macro  lowpass_8H      r0,  r1
        ext             v0.16B,     \r0\().16B, \r0\().16B, #2
        ext             v1.16B,     \r0\().16B, \r0\().16B, #3
        uaddl           v0.8H,      v0.8B,      v1.8B
        ext             v2.16B,     \r0\().16B, \r0\().16B, #1
        ext             v3.16B,     \r0\().16B, \r0\().16B, #4
        uaddl           v2.8H,      v2.8B,      v3.8B
        ext             v30.16B,    \r0\().16B, \r0\().16B, #5
        uaddl           \r0\().8H,  \r0\().8B,  v30.8B
        ext             v4.16B,     \r1\().16B, \r1\().16B, #2
        mla             \r0\().8H,  v0.8H,      v6.H[1]
        ext             v5.16B,     \r1\().16B, \r1\().16B, #3
        uaddl           v4.8H,      v4.8B,      v5.8B
        ext             v7.16B,     \r1\().16B, \r1\().16B, #1
        mls             \r0\().8H,  v2.8H,      v6.H[0]
        ext             v0.16B,     \r1\().16B, \r1\().16B, #4
        uaddl           v7.8H,      v7.8B,      v0.8B
        ext             v31.16B,    \r1\().16B, \r1\().16B, #5
        uaddl           \r1\().8H,  \r1\().8B,  v31.8B
        mla             \r1\().8H,  v4.8H,      v6.H[1]
        mls             \r1\().8H,  v7.8H,      v6.H[0]
.endm

// trashes v2-v5, v30
.macro  lowpass_8_1     r0,  r1,  d0,  narrow=1
        ext             v2.8B,     \r0\().8B, \r1\().8B, #2
        ext             v3.8B,     \r0\().8B, \r1\().8B, #3
        uaddl           v2.8H,     v2.8B,     v3.8B
        ext             v4.8B,     \r0\().8B, \r1\().8B, #1
        ext             v5.8B,     \r0\().8B, \r1\().8B, #4
        uaddl           v4.8H,     v4.8B,     v5.8B
        ext             v30.8B,    \r0\().8B, \r1\().8B, #5
        uaddl           \d0\().8H, \r0\().8B, v30.8B
        mla             \d0\().8H, v2.8H,     v6.H[1]
        mls             \d0\().8H, v4.8H,     v6.H[0]
  .if \narrow
        sqrshrun        \d0\().8B, \d0\().8H, #5
  .endif
.endm

// trashed v0-v7
.macro  lowpass_8.16    r0,  r1,  r2
        ext             v1.16B,     \r0\().16B, \r1\().16B, #4
        ext             v0.16B,     \r0\().16B, \r1\().16B, #6
        saddl           v5.4S,      v1.4H,      v0.4H
        ext             v2.16B,     \r0\().16B, \r1\().16B, #2
        saddl2          v1.4S,      v1.8H,      v0.8H
        ext             v3.16B,     \r0\().16B, \r1\().16B, #8
        saddl           v6.4S,      v2.4H,      v3.4H
        ext             \r1\().16B, \r0\().16B, \r1\().16B, #10
        saddl2          v2.4S,      v2.8H,      v3.8H
        saddl           v0.4S,      \r0\().4H,  \r1\().4H
        saddl2          v4.4S,      \r0\().8H,  \r1\().8H

        shl             v3.4S,  v5.4S,  #4
        shl             v5.4S,  v5.4S,  #2
        shl             v7.4S,  v6.4S,  #2
        add             v5.4S,  v5.4S,  v3.4S
        add             v6.4S,  v6.4S,  v7.4S

        shl             v3.4S,  v1.4S,  #4
        shl             v1.4S,  v1.4S,  #2
        shl             v7.4S,  v2.4S,  #2
        add             v1.4S,  v1.4S,  v3.4S
        add             v2.4S,  v2.4S,  v7.4S

        add             v5.4S,  v5.4S,  v0.4S
        sub             v5.4S,  v5.4S,  v6.4S

        add             v1.4S,  v1.4S,  v4.4S
        sub             v1.4S,  v1.4S,  v2.4S

        rshrn           v5.4H,  v5.4S,  #10
        rshrn2          v5.8H,  v1.4S,  #10

        sqxtun          \r2\().8B,  v5.8H
.endm

function put_h264_qpel16_h_lowpass_neon_packed
        mov             x4,  x30
        mov             x12, #16
        mov             x3,  #8
        bl              put_h264_qpel8_h_lowpass_neon
        sub             x1,  x1,  x2, lsl #4
        add             x1,  x1,  #8
        mov             x12, #16
        mov             x30, x4
        b               put_h264_qpel8_h_lowpass_neon
endfunc

.macro  h264_qpel_h_lowpass type
function \type\()_h264_qpel16_h_lowpass_neon
        mov             x13, x30
        mov             x12, #16
        bl              \type\()_h264_qpel8_h_lowpass_neon
        sub             x0,  x0,  x3, lsl #4
        sub             x1,  x1,  x2, lsl #4
        add             x0,  x0,  #8
        add             x1,  x1,  #8
        mov             x12, #16
        mov             x30, x13
endfunc

function \type\()_h264_qpel8_h_lowpass_neon
1:      ld1             {v28.8B, v29.8B}, [x1], x2
        ld1             {v16.8B, v17.8B}, [x1], x2
        subs            x12, x12, #2
        lowpass_8       v28, v29, v16, v17, v28, v16
  .ifc \type,avg
        ld1             {v2.8B},    [x0], x3
        urhadd          v28.8B, v28.8B,  v2.8B
        ld1             {v3.8B},    [x0]
        urhadd          v16.8B, v16.8B, v3.8B
        sub             x0,  x0,  x3
  .endif
        st1             {v28.8B},    [x0], x3
        st1             {v16.8B},    [x0], x3
        b.ne            1b
        ret
endfunc
.endm

        h264_qpel_h_lowpass put
        h264_qpel_h_lowpass avg

.macro  h264_qpel_h_lowpass_l2 type
function \type\()_h264_qpel16_h_lowpass_l2_neon
        mov             x13, x30
        mov             x12, #16
        bl              \type\()_h264_qpel8_h_lowpass_l2_neon
        sub             x0,  x0,  x2, lsl #4
        sub             x1,  x1,  x2, lsl #4
        sub             x3,  x3,  x2, lsl #4
        add             x0,  x0,  #8
        add             x1,  x1,  #8
        add             x3,  x3,  #8
        mov             x12, #16
        mov             x30, x13
endfunc

function \type\()_h264_qpel8_h_lowpass_l2_neon
1:      ld1             {v26.8B, v27.8B}, [x1], x2
        ld1             {v16.8B, v17.8B}, [x1], x2
        ld1             {v28.8B},     [x3], x2
        ld1             {v29.8B},     [x3], x2
        subs            x12, x12, #2
        lowpass_8       v26, v27, v16, v17, v26, v27
        urhadd          v26.8B, v26.8B, v28.8B
        urhadd          v27.8B, v27.8B, v29.8B
  .ifc \type,avg
        ld1             {v2.8B},      [x0], x2
        urhadd          v26.8B, v26.8B, v2.8B
        ld1             {v3.8B},      [x0]
        urhadd          v27.8B, v27.8B, v3.8B
        sub             x0,  x0,  x2
  .endif
        st1             {v26.8B},     [x0], x2
        st1             {v27.8B},     [x0], x2
        b.ne            1b
        ret
endfunc
.endm

        h264_qpel_h_lowpass_l2 put
        h264_qpel_h_lowpass_l2 avg

function put_h264_qpel16_v_lowpass_neon_packed
        mov             x4,  x30
        mov             x2,  #8
        bl              put_h264_qpel8_v_lowpass_neon
        sub             x1,  x1,  x3, lsl #2
        bl              put_h264_qpel8_v_lowpass_neon
        sub             x1,  x1,  x3, lsl #4
        sub             x1,  x1,  x3, lsl #2
        add             x1,  x1,  #8
        bl              put_h264_qpel8_v_lowpass_neon
        sub             x1,  x1,  x3, lsl #2
        mov             x30, x4
        b               put_h264_qpel8_v_lowpass_neon
endfunc

.macro  h264_qpel_v_lowpass type
function \type\()_h264_qpel16_v_lowpass_neon
        mov             x4,  x30
        bl              \type\()_h264_qpel8_v_lowpass_neon
        sub             x1,  x1,  x3, lsl #2
        bl              \type\()_h264_qpel8_v_lowpass_neon
        sub             x0,  x0,  x2, lsl #4
        add             x0,  x0,  #8
        sub             x1,  x1,  x3, lsl #4
        sub             x1,  x1,  x3, lsl #2
        add             x1,  x1,  #8
        bl              \type\()_h264_qpel8_v_lowpass_neon
        sub             x1,  x1,  x3, lsl #2
        mov             x30, x4
endfunc

function \type\()_h264_qpel8_v_lowpass_neon
        ld1             {v16.8B}, [x1], x3
        ld1             {v18.8B}, [x1], x3
        ld1             {v20.8B}, [x1], x3
        ld1             {v22.8B}, [x1], x3
        ld1             {v24.8B}, [x1], x3
        ld1             {v26.8B}, [x1], x3
        ld1             {v28.8B}, [x1], x3
        ld1             {v30.8B}, [x1], x3
        ld1             {v17.8B}, [x1], x3
        ld1             {v19.8B}, [x1], x3
        ld1             {v21.8B}, [x1], x3
        ld1             {v23.8B}, [x1], x3
        ld1             {v25.8B}, [x1]

        transpose_8x8B  v16, v18, v20, v22, v24, v26, v28, v30, v0,  v1
        transpose_8x8B  v17, v19, v21, v23, v25, v27, v29, v31, v0,  v1
        lowpass_8       v16, v17, v18, v19, v16, v17
        lowpass_8       v20, v21, v22, v23, v18, v19
        lowpass_8       v24, v25, v26, v27, v20, v21
        lowpass_8       v28, v29, v30, v31, v22, v23
        transpose_8x8B  v16, v17, v18, v19, v20, v21, v22, v23, v0,  v1

  .ifc \type,avg
        ld1             {v24.8B},  [x0], x2
        urhadd          v16.8B, v16.8B, v24.8B
        ld1             {v25.8B}, [x0], x2
        urhadd          v17.8B, v17.8B, v25.8B
        ld1             {v26.8B}, [x0], x2
        urhadd          v18.8B, v18.8B, v26.8B
        ld1             {v27.8B}, [x0], x2
        urhadd          v19.8B, v19.8B, v27.8B
        ld1             {v28.8B}, [x0], x2
        urhadd          v20.8B, v20.8B, v28.8B
        ld1             {v29.8B}, [x0], x2
        urhadd          v21.8B, v21.8B, v29.8B
        ld1             {v30.8B}, [x0], x2
        urhadd          v22.8B, v22.8B, v30.8B
        ld1             {v31.8B}, [x0], x2
        urhadd          v23.8B, v23.8B, v31.8B
        sub             x0,  x0,  x2,  lsl #3
  .endif

        st1             {v16.8B}, [x0], x2
        st1             {v17.8B}, [x0], x2
        st1             {v18.8B}, [x0], x2
        st1             {v19.8B}, [x0], x2
        st1             {v20.8B}, [x0], x2
        st1             {v21.8B}, [x0], x2
        st1             {v22.8B}, [x0], x2
        st1             {v23.8B}, [x0], x2

        ret
endfunc
.endm

        h264_qpel_v_lowpass put
        h264_qpel_v_lowpass avg

.macro  h264_qpel_v_lowpass_l2 type
function \type\()_h264_qpel16_v_lowpass_l2_neon
        mov             x4,  x30
        bl              \type\()_h264_qpel8_v_lowpass_l2_neon
        sub             x1,  x1,  x3, lsl #2
        bl              \type\()_h264_qpel8_v_lowpass_l2_neon
        sub             x0,  x0,  x3, lsl #4
        sub             x12, x12, x2, lsl #4
        add             x0,  x0,  #8
        add             x12, x12, #8
        sub             x1,  x1,  x3, lsl #4
        sub             x1,  x1,  x3, lsl #2
        add             x1,  x1,  #8
        bl              \type\()_h264_qpel8_v_lowpass_l2_neon
        sub             x1,  x1,  x3, lsl #2
        mov             x30, x4
endfunc

function \type\()_h264_qpel8_v_lowpass_l2_neon
        ld1             {v16.8B}, [x1], x3
        ld1             {v18.8B}, [x1], x3
        ld1             {v20.8B}, [x1], x3
        ld1             {v22.8B}, [x1], x3
        ld1             {v24.8B}, [x1], x3
        ld1             {v26.8B}, [x1], x3
        ld1             {v28.8B}, [x1], x3
        ld1             {v30.8B}, [x1], x3
        ld1             {v17.8B}, [x1], x3
        ld1             {v19.8B}, [x1], x3
        ld1             {v21.8B}, [x1], x3
        ld1             {v23.8B}, [x1], x3
        ld1             {v25.8B}, [x1]

        transpose_8x8B  v16, v18, v20, v22, v24, v26, v28, v30, v0,  v1
        transpose_8x8B  v17, v19, v21, v23, v25, v27, v29, v31, v0,  v1
        lowpass_8       v16, v17, v18, v19, v16, v17
        lowpass_8       v20, v21, v22, v23, v18, v19
        lowpass_8       v24, v25, v26, v27, v20, v21
        lowpass_8       v28, v29, v30, v31, v22, v23
        transpose_8x8B  v16, v17, v18, v19, v20, v21, v22, v23, v0,  v1

        ld1             {v24.8B},  [x12], x2
        ld1             {v25.8B},  [x12], x2
        ld1             {v26.8B},  [x12], x2
        ld1             {v27.8B},  [x12], x2
        ld1             {v28.8B},  [x12], x2
        urhadd          v16.8B, v24.8B, v16.8B
        urhadd          v17.8B, v25.8B, v17.8B
        ld1             {v29.8B},  [x12], x2
        urhadd          v18.8B, v26.8B, v18.8B
        urhadd          v19.8B, v27.8B, v19.8B
        ld1             {v30.8B}, [x12], x2
        urhadd          v20.8B, v28.8B, v20.8B
        urhadd          v21.8B, v29.8B, v21.8B
        ld1             {v31.8B}, [x12], x2
        urhadd          v22.8B, v30.8B, v22.8B
        urhadd          v23.8B, v31.8B, v23.8B

  .ifc \type,avg
        ld1             {v24.8B}, [x0], x3
        urhadd          v16.8B, v16.8B, v24.8B
        ld1             {v25.8B}, [x0], x3
        urhadd          v17.8B, v17.8B, v25.8B
        ld1             {v26.8B}, [x0], x3
        urhadd          v18.8B, v18.8B, v26.8B
        ld1             {v27.8B}, [x0], x3
        urhadd          v19.8B, v19.8B, v27.8B
        ld1             {v28.8B}, [x0], x3
        urhadd          v20.8B, v20.8B, v28.8B
        ld1             {v29.8B}, [x0], x3
        urhadd          v21.8B, v21.8B, v29.8B
        ld1             {v30.8B}, [x0], x3
        urhadd          v22.8B, v22.8B, v30.8B
        ld1             {v31.8B}, [x0], x3
        urhadd          v23.8B, v23.8B, v31.8B
        sub             x0,  x0,  x3,  lsl #3
  .endif

        st1             {v16.8B}, [x0], x3
        st1             {v17.8B}, [x0], x3
        st1             {v18.8B}, [x0], x3
        st1             {v19.8B}, [x0], x3
        st1             {v20.8B}, [x0], x3
        st1             {v21.8B}, [x0], x3
        st1             {v22.8B}, [x0], x3
        st1             {v23.8B}, [x0], x3

        ret
endfunc
.endm

        h264_qpel_v_lowpass_l2 put
        h264_qpel_v_lowpass_l2 avg

function put_h264_qpel8_hv_lowpass_neon_top
        lowpass_const   w12
        ld1             {v16.8H}, [x1], x3
        ld1             {v17.8H}, [x1], x3
        ld1             {v18.8H}, [x1], x3
        ld1             {v19.8H}, [x1], x3
        ld1             {v20.8H}, [x1], x3
        ld1             {v21.8H}, [x1], x3
        ld1             {v22.8H}, [x1], x3
        ld1             {v23.8H}, [x1], x3
        ld1             {v24.8H}, [x1], x3
        ld1             {v25.8H}, [x1], x3
        ld1             {v26.8H}, [x1], x3
        ld1             {v27.8H}, [x1], x3
        ld1             {v28.8H}, [x1]
        lowpass_8H      v16, v17
        lowpass_8H      v18, v19
        lowpass_8H      v20, v21
        lowpass_8H      v22, v23
        lowpass_8H      v24, v25
        lowpass_8H      v26, v27
        lowpass_8H      v28, v29

        transpose_8x8H  v16, v17, v18, v19, v20, v21, v22, v23, v0,  v1
        transpose_8x8H  v24, v25, v26, v27, v28, v29, v30, v31, v0,  v1

        lowpass_8.16    v16, v24, v16
        lowpass_8.16    v17, v25, v17

        lowpass_8.16    v18, v26, v18
        lowpass_8.16    v19, v27, v19

        lowpass_8.16    v20, v28, v20
        lowpass_8.16    v21, v29, v21

        lowpass_8.16    v22, v30, v22
        lowpass_8.16    v23, v31, v23

        transpose_8x8B v16, v17, v18, v19, v20, v21, v22, v23, v0,  v1

        ret
endfunc

.macro  h264_qpel8_hv_lowpass type
function \type\()_h264_qpel8_hv_lowpass_neon
        mov             x10, x30
        bl              put_h264_qpel8_hv_lowpass_neon_top
  .ifc \type,avg
        ld1             {v0.8B},      [x0], x2
        urhadd          v16.8B, v16.8B, v0.8B
        ld1             {v1.8B},      [x0], x2
        urhadd          v17.8B, v17.8B, v1.8B
        ld1             {v2.8B},      [x0], x2
        urhadd          v18.8B, v18.8B, v2.8B
        ld1             {v3.8B},      [x0], x2
        urhadd          v19.8B, v19.8B, v3.8B
        ld1             {v4.8B},      [x0], x2
        urhadd          v20.8B, v20.8B, v4.8B
        ld1             {v5.8B},      [x0], x2
        urhadd          v21.8B, v21.8B, v5.8B
        ld1             {v6.8B},      [x0], x2
        urhadd          v22.8B, v22.8B, v6.8B
        ld1             {v7.8B},      [x0], x2
        urhadd          v23.8B, v23.8B, v7.8B
        sub             x0,  x0,  x2,  lsl #3
  .endif

        st1             {v16.8B},     [x0], x2
        st1             {v17.8B},     [x0], x2
        st1             {v18.8B},     [x0], x2
        st1             {v19.8B},     [x0], x2
        st1             {v20.8B},     [x0], x2
        st1             {v21.8B},     [x0], x2
        st1             {v22.8B},     [x0], x2
        st1             {v23.8B},     [x0], x2

        ret             x10
endfunc
.endm

        h264_qpel8_hv_lowpass put
        h264_qpel8_hv_lowpass avg

.macro  h264_qpel8_hv_lowpass_l2 type
function \type\()_h264_qpel8_hv_lowpass_l2_neon
        mov             x10, x30
        bl              put_h264_qpel8_hv_lowpass_neon_top

        ld1             {v0.8B, v1.8B},  [x2], #16
        ld1             {v2.8B, v3.8B},  [x2], #16
        urhadd          v0.8B,  v0.8B,  v16.8B
        urhadd          v1.8B,  v1.8B,  v17.8B
        ld1             {v4.8B, v5.8B},  [x2], #16
        urhadd          v2.8B,  v2.8B,  v18.8B
        urhadd          v3.8B,  v3.8B,  v19.8B
        ld1             {v6.8B, v7.8B},  [x2], #16
        urhadd          v4.8B,  v4.8B,  v20.8B
        urhadd          v5.8B,  v5.8B,  v21.8B
        urhadd          v6.8B,  v6.8B,  v22.8B
        urhadd          v7.8B,  v7.8B,  v23.8B
  .ifc \type,avg
        ld1             {v16.8B},     [x0], x3
        urhadd          v0.8B,  v0.8B,  v16.8B
        ld1             {v17.8B},     [x0], x3
        urhadd          v1.8B,  v1.8B,  v17.8B
        ld1             {v18.8B},     [x0], x3
        urhadd          v2.8B,  v2.8B,  v18.8B
        ld1             {v19.8B},     [x0], x3
        urhadd          v3.8B,  v3.8B,  v19.8B
        ld1             {v20.8B},     [x0], x3
        urhadd          v4.8B,  v4.8B,  v20.8B
        ld1             {v21.8B},     [x0], x3
        urhadd          v5.8B,  v5.8B,  v21.8B
        ld1             {v22.8B},     [x0], x3
        urhadd          v6.8B,  v6.8B,  v22.8B
        ld1             {v23.8B},     [x0], x3
        urhadd          v7.8B,  v7.8B,  v23.8B
        sub             x0,  x0,  x3,  lsl #3
  .endif
        st1             {v0.8B},      [x0], x3
        st1             {v1.8B},      [x0], x3
        st1             {v2.8B},      [x0], x3
        st1             {v3.8B},      [x0], x3
        st1             {v4.8B},      [x0], x3
        st1             {v5.8B},      [x0], x3
        st1             {v6.8B},      [x0], x3
        st1             {v7.8B},      [x0], x3

        ret             x10
endfunc
.endm

        h264_qpel8_hv_lowpass_l2 put
        h264_qpel8_hv_lowpass_l2 avg

.macro  h264_qpel16_hv  type
function \type\()_h264_qpel16_hv_lowpass_neon
        mov             x13, x30
        bl              \type\()_h264_qpel8_hv_lowpass_neon
        sub             x1,  x1,  x3, lsl #2
        bl              \type\()_h264_qpel8_hv_lowpass_neon
        sub             x1,  x1,  x3, lsl #4
        sub             x1,  x1,  x3, lsl #2
        add             x1,  x1,  #8
        sub             x0,  x0,  x2, lsl #4
        add             x0,  x0,  #8
        bl              \type\()_h264_qpel8_hv_lowpass_neon
        sub             x1,  x1,  x3, lsl #2
        mov             x30, x13
        b               \type\()_h264_qpel8_hv_lowpass_neon
endfunc

function \type\()_h264_qpel16_hv_lowpass_l2_neon
        mov             x13, x30
        sub             x2,  x4,  #256
        bl              \type\()_h264_qpel8_hv_lowpass_l2_neon
        sub             x1,  x1,  x3, lsl #2
        bl              \type\()_h264_qpel8_hv_lowpass_l2_neon
        sub             x1,  x1,  x3, lsl #4
        sub             x1,  x1,  x3, lsl #2
        add             x1,  x1,  #8
        sub             x0,  x0,  x3, lsl #4
        add             x0,  x0,  #8
        bl              \type\()_h264_qpel8_hv_lowpass_l2_neon
        sub             x1,  x1,  x3, lsl #2
        mov             x30, x13
        b               \type\()_h264_qpel8_hv_lowpass_l2_neon
endfunc
.endm

        h264_qpel16_hv put
        h264_qpel16_hv avg

.macro  h264_qpel8      type
function ff_\type\()_h264_qpel8_mc10_neon, export=1
        lowpass_const   w3
        mov             x3,  x1
        sub             x1,  x1,  #2
        mov             x12, #8
        b               \type\()_h264_qpel8_h_lowpass_l2_neon
endfunc

function ff_\type\()_h264_qpel8_mc20_neon, export=1
        lowpass_const   w3
        sub             x1,  x1,  #2
        mov             x3,  x2
        mov             x12, #8
        b               \type\()_h264_qpel8_h_lowpass_neon
endfunc

function ff_\type\()_h264_qpel8_mc30_neon, export=1
        lowpass_const   w3
        add             x3,  x1,  #1
        sub             x1,  x1,  #2
        mov             x12, #8
        b               \type\()_h264_qpel8_h_lowpass_l2_neon
endfunc

function ff_\type\()_h264_qpel8_mc01_neon, export=1
        mov             x14, x30
        mov             x12, x1
\type\()_h264_qpel8_mc01:
        lowpass_const   w3
        mov             x3,  x2
        sub             x1,  x1,  x2, lsl #1
        bl              \type\()_h264_qpel8_v_lowpass_l2_neon
        ret             x14
endfunc

function ff_\type\()_h264_qpel8_mc11_neon, export=1
        mov             x14, x30
        mov             x8,  x0
        mov             x9,  x1
\type\()_h264_qpel8_mc11:
        lowpass_const   w3
        mov             x11, sp
        sub             sp,  sp,  #64
        mov             x0,  sp
        sub             x1,  x1,  #2
        mov             x3,  #8
        mov             x12, #8
        bl              put_h264_qpel8_h_lowpass_neon
        mov             x0,  x8
        mov             x3,  x2
        mov             x12, sp
        sub             x1,  x9,  x2, lsl #1
        mov             x2,  #8
        bl              \type\()_h264_qpel8_v_lowpass_l2_neon
        mov             sp,  x11
        ret             x14
endfunc

function ff_\type\()_h264_qpel8_mc21_neon, export=1
        mov             x14, x30
        mov             x8,  x0
        mov             x9,  x1
\type\()_h264_qpel8_mc21:
        lowpass_const   w3
        mov             x11, sp
        sub             sp,  sp,  #(8*8+16*12)
        sub             x1,  x1,  #2
        mov             x3,  #8
        mov             x0,  sp
        mov             x12, #8
        bl              put_h264_qpel8_h_lowpass_neon
        mov             x4,  x0
        mov             x0,  x8
        sub             x1,  x9,  x2, lsl #1
        sub             x1,  x1,  #2
        mov             x3,  x2
        sub             x2,  x4,  #64
        bl              \type\()_h264_qpel8_hv_lowpass_l2_neon
        mov             sp,  x11
        ret             x14
endfunc

function ff_\type\()_h264_qpel8_mc31_neon, export=1
        add             x1,  x1,  #1
        mov             x14, x30
        mov             x8,  x0
        mov             x9,  x1
        sub             x1,  x1,  #1
        b               \type\()_h264_qpel8_mc11
endfunc

function ff_\type\()_h264_qpel8_mc02_neon, export=1
        mov             x14, x30
        lowpass_const   w3
        sub             x1,  x1,  x2, lsl #1
        mov             x3,  x2
        bl              \type\()_h264_qpel8_v_lowpass_neon
        ret             x14
endfunc

function ff_\type\()_h264_qpel8_mc12_neon, export=1
        mov             x14, x30
        mov             x8,  x0
        mov             x9,  x1
\type\()_h264_qpel8_mc12:
        lowpass_const   w3
        mov             x11, sp
        sub             sp,  sp,  #(8*8+16*12)
        sub             x1,  x1,  x2, lsl #1
        mov             x3,  x2
        mov             x2,  #8
        mov             x0,  sp
        bl              put_h264_qpel8_v_lowpass_neon
        mov             x4,  x0
        mov             x0,  x8
        sub             x1,  x9,  x3, lsl #1
        sub             x1,  x1,  #2
        sub             x2,  x4,  #64
        bl              \type\()_h264_qpel8_hv_lowpass_l2_neon
        mov             sp,  x11
        ret             x14
endfunc

function ff_\type\()_h264_qpel8_mc22_neon, export=1
        mov             x14, x30
        mov             x11, sp
        sub             x1,  x1,  x2, lsl #1
        sub             x1,  x1,  #2
        mov             x3,  x2
        bl              \type\()_h264_qpel8_hv_lowpass_neon
        mov             sp,  x11
        ret             x14
endfunc

function ff_\type\()_h264_qpel8_mc32_neon, export=1
        mov             x14, x30
        mov             x8,  x0
        mov             x9,  x1
        add             x1,  x1,  #1
        b               \type\()_h264_qpel8_mc12
endfunc

function ff_\type\()_h264_qpel8_mc03_neon, export=1
        mov             x14, x30
        add             x12, x1,  x2
        b               \type\()_h264_qpel8_mc01
endfunc

function ff_\type\()_h264_qpel8_mc13_neon, export=1
        mov             x14, x30
        mov             x8,  x0
        mov             x9,  x1
        add             x1,  x1,  x2
        b               \type\()_h264_qpel8_mc11
endfunc

function ff_\type\()_h264_qpel8_mc23_neon, export=1
        mov             x14, x30
        mov             x8,  x0
        mov             x9,  x1
        add             x1,  x1,  x2
        b               \type\()_h264_qpel8_mc21
endfunc

function ff_\type\()_h264_qpel8_mc33_neon, export=1
        add             x1,  x1,  #1
        mov             x14, x30
        mov             x8,  x0
        mov             x9,  x1
        add             x1,  x1,  x2
        sub             x1,  x1,  #1
        b               \type\()_h264_qpel8_mc11
endfunc
.endm

        h264_qpel8 put
        h264_qpel8 avg

.macro  h264_qpel16     type
function ff_\type\()_h264_qpel16_mc10_neon, export=1
        lowpass_const   w3
        mov             x3,  x1
        sub             x1,  x1,  #2
        b               \type\()_h264_qpel16_h_lowpass_l2_neon
endfunc

function ff_\type\()_h264_qpel16_mc20_neon, export=1
        lowpass_const   w3
        sub             x1,  x1,  #2
        mov             x3,  x2
        b               \type\()_h264_qpel16_h_lowpass_neon
endfunc

function ff_\type\()_h264_qpel16_mc30_neon, export=1
        lowpass_const   w3
        add             x3,  x1,  #1
        sub             x1,  x1,  #2
        b               \type\()_h264_qpel16_h_lowpass_l2_neon
endfunc

function ff_\type\()_h264_qpel16_mc01_neon, export=1
        mov             x14, x30
        mov             x12, x1
\type\()_h264_qpel16_mc01:
        lowpass_const   w3
        mov             x3,  x2
        sub             x1,  x1,  x2, lsl #1
        bl              \type\()_h264_qpel16_v_lowpass_l2_neon
        ret             x14
endfunc

function ff_\type\()_h264_qpel16_mc11_neon, export=1
        mov             x14, x30
        mov             x8,  x0
        mov             x9,  x1
\type\()_h264_qpel16_mc11:
        lowpass_const   w3
        mov             x11, sp
        sub             sp,  sp,  #256
        mov             x0,  sp
        sub             x1,  x1,  #2
        mov             x3,  #16
        bl              put_h264_qpel16_h_lowpass_neon
        mov             x0,  x8
        mov             x3,  x2
        mov             x12, sp
        sub             x1,  x9,  x2, lsl #1
        mov             x2,  #16
        bl              \type\()_h264_qpel16_v_lowpass_l2_neon
        mov             sp,  x11
        ret             x14
endfunc

function ff_\type\()_h264_qpel16_mc21_neon, export=1
        mov             x14, x30
        mov             x8,  x0
        mov             x9,  x1
\type\()_h264_qpel16_mc21:
        lowpass_const   w3
        mov             x11, sp
        sub             sp,  sp,  #(16*16+16*12)
        sub             x1,  x1,  #2
        mov             x0,  sp
        bl              put_h264_qpel16_h_lowpass_neon_packed
        mov             x4,  x0
        mov             x0,  x8
        sub             x1,  x9,  x2, lsl #1
        sub             x1,  x1,  #2
        mov             x3,  x2
        bl              \type\()_h264_qpel16_hv_lowpass_l2_neon
        mov             sp,  x11
        ret             x14
endfunc

function ff_\type\()_h264_qpel16_mc31_neon, export=1
        add             x1,  x1,  #1
        mov             x14, x30
        mov             x8,  x0
        mov             x9,  x1
        sub             x1,  x1,  #1
        b               \type\()_h264_qpel16_mc11
endfunc

function ff_\type\()_h264_qpel16_mc02_neon, export=1
        mov             x14, x30
        lowpass_const   w3
        sub             x1,  x1,  x2, lsl #1
        mov             x3,  x2
        bl              \type\()_h264_qpel16_v_lowpass_neon
        ret             x14
endfunc

function ff_\type\()_h264_qpel16_mc12_neon, export=1
        mov             x14, x30
        mov             x8,  x0
        mov             x9,  x1
\type\()_h264_qpel16_mc12:
        lowpass_const   w3
        mov             x11, sp
        sub             sp,  sp,  #(16*16+16*12)
        sub             x1,  x1,  x2, lsl #1
        mov             x0,  sp
        mov             x3,  x2
        bl              put_h264_qpel16_v_lowpass_neon_packed
        mov             x4,  x0
        mov             x0,  x8
        sub             x1,  x9,  x3, lsl #1
        sub             x1,  x1,  #2
        mov             x2,  x3
        bl              \type\()_h264_qpel16_hv_lowpass_l2_neon
        mov             sp,  x11
        ret             x14
endfunc

function ff_\type\()_h264_qpel16_mc22_neon, export=1
        mov             x14, x30
        lowpass_const   w3
        mov             x11, sp
        sub             x1,  x1,  x2, lsl #1
        sub             x1,  x1,  #2
        mov             x3,  x2
        bl              \type\()_h264_qpel16_hv_lowpass_neon
        mov             sp,  x11 // restore stack
        ret             x14
endfunc

function ff_\type\()_h264_qpel16_mc32_neon, export=1
        mov             x14, x30
        mov             x8,  x0
        mov             x9,  x1
        add             x1,  x1,  #1
        b               \type\()_h264_qpel16_mc12
endfunc

function ff_\type\()_h264_qpel16_mc03_neon, export=1
        mov             x14, x30
        add             x12, x1,  x2
        b               \type\()_h264_qpel16_mc01
endfunc

function ff_\type\()_h264_qpel16_mc13_neon, export=1
        mov             x14, x30
        mov             x8,  x0
        mov             x9,  x1
        add             x1,  x1,  x2
        b               \type\()_h264_qpel16_mc11
endfunc

function ff_\type\()_h264_qpel16_mc23_neon, export=1
        mov             x14, x30
        mov             x8,  x0
        mov             x9,  x1
        add             x1,  x1,  x2
        b               \type\()_h264_qpel16_mc21
endfunc

function ff_\type\()_h264_qpel16_mc33_neon, export=1
        add             x1,  x1,  #1
        mov             x14, x30
        mov             x8,  x0
        mov             x9,  x1
        add             x1,  x1,  x2
        sub             x1,  x1,  #1
        b               \type\()_h264_qpel16_mc11
endfunc
.endm

        h264_qpel16 put
        h264_qpel16 avg