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
|
//******************************************************************************
//*
//* Copyright (C) 2015 The Android Open Source Project
//*
//* Licensed under the Apache License, Version 2.0 (the "License");
//* you may not use this file except in compliance with the License.
//* You may obtain a copy of the License at:
//*
//* http://www.apache.org/licenses/LICENSE-2.0
//*
//* Unless required by applicable law or agreed to in writing, software
//* distributed under the License is distributed on an "AS IS" BASIS,
//* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
//* See the License for the specific language governing permissions and
//* limitations under the License.
//*
//*****************************************************************************
//* Originally developed and contributed by Ittiam Systems Pvt. Ltd, Bangalore
//*/
///**
///*******************************************************************************
// * //file
// * ih264_iquant_itrans_recon_a9.s
// *
// * //brief
// * Contains function definitions for single stage inverse transform
// *
// * //author
// * Parthiban V
// * Mohit
// * Harinarayanaan
// *
// * //par List of Functions:
// * - ih264_iquant_itrans_recon_4x4_av8()
// * - ih264_iquant_itrans_recon_8x8_av8()
// * - ih264_iquant_itrans_recon_chroma_4x4_av8()
// *
// * //remarks
// * None
// *
// *******************************************************************************
.text
.p2align 2
.include "ih264_neon_macros.s"
///*
// *******************************************************************************
// *
// * //brief
// * This function performs inverse quant and Inverse transform type Ci4 for 4*4 block
// *
// * //par Description:
// * Performs inverse transform Ci4 and adds the residue to get the
// * reconstructed block
// *
// * //param[in] pi2_src
// * Input 4x4 coefficients
// *
// * //param[in] pu1_pred
// * Prediction 4x4 block
// *
// * //param[out] pu1_out
// * Output 4x4 block
// *
// * //param[in] u4_qp_div_6
// * QP
// *
// * //param[in] pu2_weigh_mat
// * Pointer to weight matrix
// *
// * //param[in] pred_strd,
// * Prediction stride
// *
// * //param[in] out_strd
// * Output Stride
// *
// *//param[in] pi2_tmp
// * temporary buffer of size 1*16
// *
// * //param[in] pu2_iscal_mat
// * Pointer to the inverse quantization matrix
// *
// * //returns Void
// *
// * //remarks
// * None
// *
// *******************************************************************************
// */
//void ih264_iquant_itrans_recon_4x4(WORD16 *pi2_src,
// UWORD8 *pu1_pred,
// UWORD8 *pu1_out,
// WORD32 pred_strd,
// WORD32 out_strd,
// const UWORD16 *pu2_iscal_mat,
// const UWORD16 *pu2_weigh_mat,
// UWORD32 u4_qp_div_6,
// WORD32 *pi4_tmp,
// WORD32 iq_start_idx
// WORD16 *pi2_dc_ld_addr)
//**************Variables Vs Registers*****************************************
//x0 => *pi2_src
//x1 => *pu1_pred
//x2 => *pu1_out
//x3 => pred_strd
//x4 => out_strd
//x5 => *pu2_iscal_mat
//x6 => *pu2_weigh_mat
//x7 => u4_qp_div_6
// => pi4_tmp
// => iq_start_idx
// => pi2_dc_ld_addr
//Only one shift is done in horizontal inverse because,
//if u4_qp_div_6 is lesser than 4 then shift value will be neagative and do negative left shift, in this case rnd_factor has value
//if u4_qp_div_6 is greater than 4 then shift value will be positive and do left shift, here rnd_factor is 0
.global ih264_iquant_itrans_recon_4x4_av8
ih264_iquant_itrans_recon_4x4_av8:
push_v_regs
dup v30.4s, w7 //Populate the u4_qp_div_6 in Q15
ldr w8, [sp, #72] //Loads iq_start_idx
sxtw x8, w8
ldr x10, [sp, #80] //Load alternate dc address
subs x8, x8, #1 // if x8 == 1 => intra case , so result of subtraction is zero and z flag is set
//=======================DEQUANT FROM HERE===================================
ld4 {v20.4h - v23.4h}, [x5] // load pu2_iscal_mat[i], i =0..15
ld4 {v26.4h - v29.4h}, [x6] // pu2_weigh_mat[i], i =0..15
ld4 {v16.4h - v19.4h}, [x0] // pi2_src_tmp[i], i =0..15
mul v20.4h, v20.4h, v26.4h // x[i]=(scale[i] * dequant[i]) where i = 0..3
mul v21.4h, v21.4h, v27.4h // x[i]=(scale[i] * dequant[i]) where i = 4..7
mul v22.4h, v22.4h, v28.4h // x[i]=(scale[i] * dequant[i]) where i = 8..11
mul v23.4h, v23.4h, v29.4h // x[i]=(scale[i] * dequant[i]) where i = 12..14
smull v0.4s, v16.4h, v20.4h // q0 = p[i] = (x[i] * trns_coeff[i]) where i = 0..3
smull v2.4s, v17.4h, v21.4h // q1 = p[i] = (x[i] * trns_coeff[i]) where i = 4..7
smull v4.4s, v18.4h, v22.4h // q2 = p[i] = (x[i] * trns_coeff[i]) where i = 8..11
smull v6.4s, v19.4h, v23.4h // q3 = p[i] = (x[i] * trns_coeff[i]) where i = 12..15
sshl v0.4s, v0.4s, v30.4s // q0 = q[i] = (p[i] << (qp/6)) where i = 0..3
sshl v2.4s, v2.4s, v30.4s // q1 = q[i] = (p[i] << (qp/6)) where i = 4..7
sshl v4.4s, v4.4s, v30.4s // q2 = q[i] = (p[i] << (qp/6)) where i = 8..11
sshl v6.4s, v6.4s, v30.4s // q3 = q[i] = (p[i] << (qp/6)) where i = 12..15
sqrshrn v0.4h, v0.4s, #0x4 // d0 = c[i] = ((q[i] + 32) >> 4) where i = 0..3
sqrshrn v1.4h, v2.4s, #0x4 // d1 = c[i] = ((q[i] + 32) >> 4) where i = 4..7
sqrshrn v2.4h, v4.4s, #0x4 // d2 = c[i] = ((q[i] + 32) >> 4) where i = 8..11
sqrshrn v3.4h, v6.4s, #0x4 // d3 = c[i] = ((q[i] + 32) >> 4) where i = 12..15
bne skip_loading_luma_dc_src
ld1 {v0.h}[0], [x10] // loads signed halfword pi2_dc_ld_addr[0], if x8==1
skip_loading_luma_dc_src:
//========= PROCESS IDCT FROM HERE =======
//Steps for Stage 1:
//------------------
ld1 {v30.s}[0], [x1], x3 // i row load pu1_pred buffer
sshr v8.4h, v1.4h, #1 // d1>>1
sshr v9.4h, v3.4h, #1 // d3>>1
add v4.4h, v0.4h, v2.4h // x0 = d0 + d2//
sub v5.4h, v0.4h, v2.4h // x1 = d0 - d2//
sub v6.4h, v8.4h, v3.4h // x2 = (d1 >> 1) - d3//
add v7.4h, v1.4h, v9.4h // x3 = d1 + (d3 >> 1)//
ld1 {v30.s}[1], [x1], x3 // ii row load pu1_pred buffer
add v10.4h, v4.4h , v7.4h // x0+x3
add v11.4h, v5.4h , v6.4h // x1+x2
sub v12.4h, v5.4h , v6.4h // x1-x2
sub v13.4h, v4.4h , v7.4h
ld1 {v31.s}[0], [x1], x3 // iii row load pu1_pred buf
//Steps for Stage 2:
//transopose
trn1 v4.4h, v10.4h, v11.4h
trn2 v5.4h, v10.4h, v11.4h
trn1 v6.4h, v12.4h, v13.4h
trn2 v7.4h, v12.4h, v13.4h
trn1 v10.2s, v4.2s, v6.2s // 0
trn1 v11.2s, v5.2s, v7.2s // 8
trn2 v12.2s, v4.2s, v6.2s // 4
trn2 v13.2s, v5.2s, v7.2s
//end transpose
sshr v18.4h, v11.4h, #1 // q0>>1
sshr v19.4h, v13.4h, #1 // q1>>1
add v14.4h, v10.4h, v12.4h // x0 = q0 + q2//
sub v15.4h, v10.4h, v12.4h // x1 = q0 - q2//
sub v16.4h, v18.4h, v13.4h // x2 = (q1 >> 1) - q3//
add v17.4h, v11.4h, v19.4h // x3 = q1+ (q3 >> 3)//
ld1 {v31.s}[1], [x1], x3 // iv row load pu1_pred buffer
add v20.4h, v14.4h, v17.4h // x0 + x3
add v21.4h, v15.4h, v16.4h // x1 + x2
sub v22.4h, v15.4h, v16.4h // x1 - x2
sub v23.4h, v14.4h, v17.4h // x0 - x3
mov v20.d[1], v21.d[0]
mov v22.d[1], v23.d[0]
srshr v20.8h, v20.8h, #6
srshr v22.8h, v22.8h, #6
uaddw v20.8h, v20.8h , v30.8b
uaddw v22.8h, v22.8h , v31.8b
sqxtun v0.8b, v20.8h
sqxtun v1.8b, v22.8h
st1 {v0.s}[0], [x2], x4 //i row store the value
st1 {v0.s}[1], [x2], x4 //ii row store the value
st1 {v1.s}[0], [x2], x4 //iii row store the value
st1 {v1.s}[1], [x2] //iv row store the value
pop_v_regs
ret
///**
// *******************************************************************************
// *
// * @brief
// * This function performs inverse quant and Inverse transform type Ci4 for 4*4 block
// *
// * @par Description:
// * Performs inverse transform Ci4 and adds the residue to get the
// * reconstructed block
// *
// * @param[in] pi2_src
// * Input 4x4 coefficients
// *
// * @param[in] pu1_pred
// * Prediction 4x4 block
// *
// * @param[out] pu1_out
// * Output 4x4 block
// *
// * @param[in] u4_qp_div_6
// * QP
// *
// * @param[in] pu2_weigh_mat
// * Pointer to weight matrix
// *
// * @param[in] pred_strd,
// * Prediction stride
// *
// * @param[in] out_strd
// * Output Stride
// *
// *@param[in] pi2_tmp
// * temporary buffer of size 1*16
// *
// * @param[in] pu2_iscal_mat
// * Pointer to the inverse quantization matrix
// *
// * @returns Void
// *
// * @remarks
// * None
// *
// *******************************************************************************
// */
//void ih264_iquant_itrans_recon_chroma_4x4(WORD16 *pi2_src,
// UWORD8 *pu1_pred,
// UWORD8 *pu1_out,
// WORD32 pred_strd,
// WORD32 out_strd,
// const UWORD16 *pu2_iscal_mat,
// const UWORD16 *pu2_weigh_mat,
// UWORD32 u4_qp_div_6,
// WORD32 *pi4_tmp
// WORD16 *pi2_dc_src)
//**************Variables Vs Registers*****************************************
//x0 => *pi2_src
//x1 => *pu1_pred
//x2 => *pu1_out
//x3 => pred_strd
//x4 => out_strd
//x5 => *pu2_iscal_mat
//x6 => *pu2_weigh_mat
//x7 => u4_qp_div_6
//sp => pi4_tmp
//sp#8 => *pi2_dc_src
.global ih264_iquant_itrans_recon_chroma_4x4_av8
ih264_iquant_itrans_recon_chroma_4x4_av8:
//VLD4.S16 is used because the pointer is incremented by SUB_BLK_WIDTH_4x4
//If the macro value changes need to change the instruction according to it.
//Only one shift is done in horizontal inverse because,
//if u4_qp_div_6 is lesser than 4 then shift value will be neagative and do negative left shift, in this case rnd_factor has value
//if u4_qp_div_6 is greater than 4 then shift value will be positive and do left shift, here rnd_factor is 0
//at the end of the fucntion, we could have moved 64 bits into heigher 64 bits of register and done further processing
//but it seem to give only reduce the number of instruction by 1. [Since a15 we saw add and sub to be very high throughput
//all instructions were taken as equal
//reduce sp by 64
push_v_regs
dup v30.4s, w7 //Populate the u4_qp_div_6 in Q15
//was at sp + 8, hence now at sp+64+8 = sp+72
ldr x10, [sp, #72] //Load alternate dc address
//=======================DEQUANT FROM HERE===================================
ld4 {v20.4h - v23.4h}, [x5] // load pu2_iscal_mat[i], i =0..15
ld4 {v26.4h - v29.4h}, [x6] // pu2_weigh_mat[i], i =0..15
ld4 {v16.4h - v19.4h}, [x0] // pi2_src_tmp[i], i =0..15
mul v20.4h, v20.4h, v26.4h // x[i]=(scale[i] * dequant[i]) where i = 0..3
mul v21.4h, v21.4h, v27.4h // x[i]=(scale[i] * dequant[i]) where i = 4..7
mul v22.4h, v22.4h, v28.4h // x[i]=(scale[i] * dequant[i]) where i = 8..11
mul v23.4h, v23.4h, v29.4h // x[i]=(scale[i] * dequant[i]) where i = 12..14
smull v0.4s, v16.4h, v20.4h // q0 = p[i] = (x[i] * trns_coeff[i]) where i = 0..3
smull v2.4s, v17.4h, v21.4h // q1 = p[i] = (x[i] * trns_coeff[i]) where i = 4..7
smull v4.4s, v18.4h, v22.4h // q2 = p[i] = (x[i] * trns_coeff[i]) where i = 8..11
smull v6.4s, v19.4h, v23.4h // q3 = p[i] = (x[i] * trns_coeff[i]) where i = 12..15
sshl v0.4s, v0.4s, v30.4s // q0 = q[i] = (p[i] << (qp/6)) where i = 0..3
sshl v2.4s, v2.4s, v30.4s // q1 = q[i] = (p[i] << (qp/6)) where i = 4..7
sshl v4.4s, v4.4s, v30.4s // q2 = q[i] = (p[i] << (qp/6)) where i = 8..11
sshl v6.4s, v6.4s, v30.4s // q3 = q[i] = (p[i] << (qp/6)) where i = 12..15
sqrshrn v0.4h, v0.4s, #0x4 // d0 = c[i] = ((q[i] + 32) >> 4) where i = 0..3
sqrshrn v1.4h, v2.4s, #0x4 // d1 = c[i] = ((q[i] + 32) >> 4) where i = 4..7
sqrshrn v2.4h, v4.4s, #0x4 // d2 = c[i] = ((q[i] + 32) >> 4) where i = 8..11
sqrshrn v3.4h, v6.4s, #0x4 // d3 = c[i] = ((q[i] + 32) >> 4) where i = 12..15
ld1 {v0.h}[0], [x10] // loads signed halfword pi2_dc_src[0]
//========= PROCESS IDCT FROM HERE =======
//Steps for Stage 1:
//------------------
sshr v8.4h, v1.4h, #1 // d1>>1
sshr v9.4h, v3.4h, #1 // d3>>1
add v4.4h, v0.4h, v2.4h // x0 = d0 + d2//
sub v5.4h, v0.4h, v2.4h // x1 = d0 - d2//
sub v6.4h, v8.4h, v3.4h // x2 = (d1 >> 1) - d3//
add v7.4h, v1.4h, v9.4h // x3 = d1 + (d3 >> 1)//
add v10.4h, v4.4h , v7.4h // x0+x3
add v11.4h, v5.4h , v6.4h // x1+x2
sub v12.4h, v5.4h , v6.4h // x1-x2
sub v13.4h, v4.4h , v7.4h
ld1 {v26.8b}, [x1], x3 // i row load pu1_pred buffer
ld1 {v27.8b}, [x1], x3 // ii row load pu1_pred buffer
ld1 {v28.8b}, [x1], x3 // iii row load pu1_pred buf
ld1 {v29.8b}, [x1], x3 // iv row load pu1_pred buffer
//Steps for Stage 2:
//transopose
trn1 v4.4h, v10.4h, v11.4h
trn2 v5.4h, v10.4h, v11.4h
trn1 v6.4h, v12.4h, v13.4h
trn2 v7.4h, v12.4h, v13.4h
trn1 v10.2s, v4.2s, v6.2s // 0
trn1 v11.2s, v5.2s, v7.2s // 8
trn2 v12.2s, v4.2s, v6.2s // 4
trn2 v13.2s, v5.2s, v7.2s
//end transpose
sshr v18.4h, v11.4h, #1 // q0>>1
sshr v19.4h, v13.4h, #1 // q1>>1
add v14.4h, v10.4h, v12.4h // x0 = q0 + q2//
sub v15.4h, v10.4h, v12.4h // x1 = q0 - q2//
sub v16.4h, v18.4h, v13.4h // x2 = (q1 >> 1) - q3//
add v17.4h, v11.4h, v19.4h // x3 = q1+ (q3 >> 3)//
//Backup the output addr
mov x0, x2
//load outpt buufer for interleaving
ld1 {v10.8b}, [x2], x4
ld1 {v11.8b}, [x2], x4
ld1 {v12.8b}, [x2], x4
ld1 {v13.8b}, [x2]
add v20.4h, v14.4h, v17.4h // x0 + x3
add v21.4h, v15.4h, v16.4h // x1 + x2
sub v22.4h, v15.4h, v16.4h // x1 - x2
sub v23.4h, v14.4h, v17.4h // x0 - x3
srshr v20.4h, v20.4h, #6
srshr v21.4h, v21.4h, #6
srshr v22.4h, v22.4h, #6
srshr v23.4h, v23.4h, #6
//nop v30.8b //dummy for deinterleaving
movi v31.4h, #0x00ff //mask for interleaving [copy lower 8 bits]
//Extract u/v plane from interleaved data
uzp1 v26.8b, v26.8b, v30.8b
uzp1 v27.8b, v27.8b, v30.8b
uzp1 v28.8b, v28.8b, v30.8b
uzp1 v29.8b, v29.8b, v30.8b
uaddw v20.8h, v20.8h, v26.8b
uaddw v21.8h, v21.8h, v27.8b
uaddw v22.8h, v22.8h, v28.8b
uaddw v23.8h, v23.8h, v29.8b
sqxtun v0.8b, v20.8h
sqxtun v1.8b, v21.8h
sqxtun v2.8b, v22.8h
sqxtun v3.8b, v23.8h
//long the output so that we have 0 at msb and value at lsb
uxtl v6.8h, v0.8b
uxtl v7.8h, v1.8b
uxtl v8.8h, v2.8b
uxtl v9.8h, v3.8b
//select lsbs from proceesd data and msbs from pu1_out loaded data
bit v10.8b, v6.8b, v31.8b
bit v11.8b, v7.8b, v31.8b
bit v12.8b, v8.8b, v31.8b
bit v13.8b, v9.8b, v31.8b
//store the interleaved result
st1 {v10.8b}, [x0], x4
st1 {v11.8b}, [x0], x4
st1 {v12.8b}, [x0], x4
st1 {v13.8b}, [x0]
pop_v_regs
ret
///*
// *******************************************************************************
// *
// * //brief
// * This function performs inverse quant and Inverse transform type Ci4 for 8*8 block
// *
// * //par Description:
// * Performs inverse transform Ci8 and adds the residue to get the
// * reconstructed block
// *
// * //param[in] pi2_src
// * Input 4x4 coefficients
// *
// * //param[in] pu1_pred
// * Prediction 4x4 block
// *
// * //param[out] pu1_out
// * Output 4x4 block
// *
// * //param[in] u4_qp_div_6
// * QP
// *
// * //param[in] pu2_weigh_mat
// * Pointer to weight matrix
// *
// * //param[in] pred_strd,
// * Prediction stride
// *
// * //param[in] out_strd
// * Output Stride
// *
// *//param[in] pi2_tmp
// * temporary buffer of size 1*64
// *
// * //param[in] pu2_iscal_mat
// * Pointer to the inverse quantization matrix
// *
// * //returns Void
// *
// * //remarks
// * None
// *
// *******************************************************************************
// */
//void ih264_iquant_itrans_recon_8x8(WORD16 *pi2_src,
// UWORD8 *pu1_pred,
// UWORD8 *pu1_out,
// WORD32 pred_strd,
// WORD32 out_strd,
// const UWORD16 *pu2_iscal_mat,
// const UWORD16 *pu2_weigh_mat,
// UWORD32 u4_qp_div_6,
// WORD32 *pi4_tmp,
// WORD32 iq_start_idx
// WORD16 *pi2_dc_ld_addr)
//**************Variables Vs Registers*****************************************
//x0 => *pi2_src
//x1 => *pu1_pred
//x2 => *pu1_out
//x3 => pred_strd
//x4 => out_strd
//x5 => *pu2_iscal_mat
//x6 => *pu2_weigh_mat
//x7 => u4_qp_div_6
//NOT USED => pi4_tmp
//NOT USED => iq_start_idx
//NOT USED => pi2_dc_ld_addr
.global ih264_iquant_itrans_recon_8x8_av8
ih264_iquant_itrans_recon_8x8_av8:
push_v_regs
ld1 {v8.8h -v11.8h}, [x5], #64
ld1 {v12.8h-v15.8h}, [x5]
ld1 {v16.8h -v19.8h}, [x6], #64
ld1 {v20.8h -v23.8h}, [x6]
mov x8, #16
ld1 {v0.8h}, [x0], x8
ld1 {v1.8h}, [x0], x8
ld1 {v2.8h}, [x0], x8
ld1 {v3.8h}, [x0], x8
ld1 {v4.8h}, [x0], x8
ld1 {v5.8h}, [x0], x8
ld1 {v6.8h}, [x0], x8
ld1 {v7.8h}, [x0]
mul v8.8h, v8.8h, v16.8h
mul v9.8h, v9.8h, v17.8h
mul v10.8h, v10.8h, v18.8h
mul v11.8h, v11.8h, v19.8h
mul v12.8h, v12.8h, v20.8h
mul v13.8h, v13.8h, v21.8h
mul v14.8h, v14.8h, v22.8h
mul v15.8h, v15.8h, v23.8h
smull v16.4s, v0.4h, v8.4h
smull2 v17.4s, v0.8h, v8.8h
smull v18.4s, v1.4h, v9.4h
smull2 v19.4s, v1.8h, v9.8h
smull v20.4s, v2.4h, v10.4h
smull2 v21.4s, v2.8h, v10.8h
smull v22.4s, v3.4h, v11.4h
smull2 v23.4s, v3.8h, v11.8h
smull v24.4s, v4.4h, v12.4h
smull2 v25.4s, v4.8h, v12.8h
smull v26.4s, v5.4h, v13.4h
smull2 v27.4s, v5.8h, v13.8h
smull v28.4s, v6.4h, v14.4h
smull2 v29.4s, v6.8h, v14.8h
smull v30.4s, v7.4h, v15.4h
smull2 v31.4s, v7.8h, v15.8h
dup v0.4s, w7
sshl v16.4s, v16.4s, v0.4s
sshl v17.4s, v17.4s, v0.4s
sshl v18.4s, v18.4s, v0.4s
sshl v19.4s, v19.4s, v0.4s
sshl v20.4s, v20.4s, v0.4s
sshl v21.4s, v21.4s, v0.4s
sshl v22.4s, v22.4s, v0.4s
sshl v23.4s, v23.4s, v0.4s
sshl v24.4s, v24.4s, v0.4s
sshl v25.4s, v25.4s, v0.4s
sshl v26.4s, v26.4s, v0.4s
sshl v27.4s, v27.4s, v0.4s
sshl v28.4s, v28.4s, v0.4s
sshl v29.4s, v29.4s, v0.4s
sshl v30.4s, v30.4s, v0.4s
sshl v31.4s, v31.4s, v0.4s
sqrshrn v0.4h, v16.4s, #6
sqrshrn2 v0.8h, v17.4s, #6
sqrshrn v1.4h, v18.4s, #6
sqrshrn2 v1.8h, v19.4s, #6
sqrshrn v2.4h, v20.4s, #6
sqrshrn2 v2.8h, v21.4s, #6
sqrshrn v3.4h, v22.4s, #6
sqrshrn2 v3.8h, v23.4s, #6
sqrshrn v4.4h, v24.4s, #6
sqrshrn2 v4.8h, v25.4s, #6
sqrshrn v5.4h, v26.4s, #6
sqrshrn2 v5.8h, v27.4s, #6
sqrshrn v6.4h, v28.4s, #6
sqrshrn2 v6.8h, v29.4s, #6
sqrshrn v7.4h, v30.4s, #6
sqrshrn2 v7.8h, v31.4s, #6
//loop counter
mov x8, #2
//1x8 transofORM
trans_1x8_1d:
//transpose 8x8
trn1 v8.8h, v0.8h, v1.8h
trn2 v9.8h, v0.8h, v1.8h
trn1 v10.8h, v2.8h, v3.8h
trn2 v11.8h, v2.8h, v3.8h
trn1 v12.8h, v4.8h, v5.8h
trn2 v13.8h, v4.8h, v5.8h
trn1 v14.8h, v6.8h, v7.8h
trn2 v15.8h, v6.8h, v7.8h
trn1 v0.4s, v8.4s, v10.4s
trn2 v2.4s, v8.4s, v10.4s
trn1 v1.4s, v9.4s, v11.4s
trn2 v3.4s, v9.4s, v11.4s
trn1 v4.4s, v12.4s, v14.4s
trn2 v6.4s, v12.4s, v14.4s
trn1 v5.4s, v13.4s, v15.4s
trn2 v7.4s, v13.4s, v15.4s
trn1 v8.2d, v0.2d, v4.2d //0
trn2 v12.2d, v0.2d, v4.2d //1
trn1 v9.2d, v1.2d, v5.2d //2
trn2 v13.2d, v1.2d, v5.2d //3
trn1 v10.2d, v2.2d, v6.2d //4
trn2 v14.2d, v2.2d, v6.2d //5
trn1 v11.2d, v3.2d, v7.2d //6
trn2 v15.2d, v3.2d, v7.2d //7
// 1 3 5 6 7
sshr v16.8h, v9.8h, #1 //(pi2_tmp_ptr[1] >> 1)
sshr v17.8h, v10.8h, #1 //(pi2_tmp_ptr[2] >> 1)
sshr v18.8h, v11.8h, #1 //(pi2_tmp_ptr[3] >> 1)
sshr v19.8h, v13.8h, #1 //(pi2_tmp_ptr[5] >> 1)
sshr v20.8h, v14.8h, #1 //(pi2_tmp_ptr[6] >> 1)
sshr v21.8h, v15.8h, #1 //(pi2_tmp_ptr[7] >> 1)
add v0.8h, v8.8h, v12.8h // i_y0 = (pi2_tmp_ptr[0] + pi2_tmp_ptr[4] );
sub v2.8h, v8.8h, v12.8h // i_y2 = (pi2_tmp_ptr[0] - pi2_tmp_ptr[4] );
sub v4.8h, v17.8h, v14.8h //i_y4 = ((pi2_tmp_ptr[2] >> 1) - pi2_tmp_ptr[6] );
add v6.8h, v10.8h, v20.8h //i_y6 = (pi2_tmp_ptr[2] + (pi2_tmp_ptr[6] >> 1));
//-w3 + w5
ssubl v22.4s, v13.4h, v11.4h
ssubl2 v23.4s, v13.8h, v11.8h
//w3 + w5
saddl v24.4s, v13.4h, v11.4h
saddl2 v25.4s, v13.8h, v11.8h
//-w1 + w7
ssubl v26.4s, v15.4h, v9.4h
ssubl2 v27.4s, v15.8h, v9.8h
//w1 + w7
saddl v28.4s, v15.4h, v9.4h
saddl2 v29.4s, v15.8h, v9.8h
//-w3 + w5 - w7
ssubw v22.4s, v22.4s, v15.4h
ssubw2 v23.4s, v23.4s, v15.8h
//w3 + w5 + w1
saddw v24.4s, v24.4s, v9.4h
saddw2 v25.4s, v25.4s, v9.8h
//-w1 + w7 + w5
saddw v26.4s, v26.4s, v13.4h
saddw2 v27.4s, v27.4s, v13.8h
//w1 + w7 - w3
ssubw v28.4s, v28.4s, v11.4h
ssubw2 v29.4s, v29.4s, v11.8h
//-w3 + w5 - w7 - (w7 >> 1)
ssubw v22.4s, v22.4s, v21.4h
ssubw2 v23.4s, v23.4s, v21.8h
//w3 + w5 + w1 + (w1 >> 1)
saddw v24.4s, v24.4s, v16.4h
saddw2 v25.4s, v25.4s, v16.8h
//-w1 + w7 + w5 + (w5 >> 1)
saddw v26.4s, v26.4s, v19.4h
saddw2 v27.4s, v27.4s, v19.8h
//w1 + w7 - w3 - (w3 >> 1)
ssubw v28.4s, v28.4s, v18.4h
ssubw2 v29.4s, v29.4s, v18.8h
xtn v1.4h, v22.4s
xtn2 v1.8h, v23.4s
xtn v3.4h, v28.4s
xtn2 v3.8h, v29.4s
xtn v5.4h, v26.4s
xtn2 v5.8h, v27.4s
xtn v7.4h, v24.4s
xtn2 v7.8h, v25.4s
sshr v16.8h, v1.8h, #2 //(y1 >> 2)
sshr v17.8h, v3.8h, #2 //(y3 >> 2)
sshr v18.8h, v5.8h, #2 //(y5 >> 2)
sshr v19.8h, v7.8h, #2 //(y7 >> 2)
add v8.8h, v0.8h, v6.8h
add v9.8h, v1.8h, v19.8h
add v10.8h, v2.8h, v4.8h
add v11.8h, v3.8h, v18.8h
sub v12.8h, v2.8h, v4.8h
sub v13.8h, v17.8h, v5.8h
sub v14.8h, v0.8h, v6.8h
sub v15.8h, v7.8h, v16.8h
add v0.8h, v8.8h, v15.8h
add v1.8h, v10.8h, v13.8h
add v2.8h, v12.8h, v11.8h
add v3.8h, v14.8h, v9.8h
sub v4.8h, v14.8h, v9.8h
sub v5.8h, v12.8h, v11.8h
sub v6.8h, v10.8h, v13.8h
sub v7.8h, v8.8h, v15.8h
subs x8, x8, #1
bne trans_1x8_1d
ld1 {v22.8b}, [x1], x3
ld1 {v23.8b}, [x1], x3
ld1 {v24.8b}, [x1], x3
ld1 {v25.8b}, [x1], x3
ld1 {v26.8b}, [x1], x3
ld1 {v27.8b}, [x1], x3
ld1 {v28.8b}, [x1], x3
ld1 {v29.8b}, [x1]
srshr v0.8h, v0.8h, #6
srshr v1.8h, v1.8h, #6
srshr v2.8h, v2.8h, #6
srshr v3.8h, v3.8h, #6
srshr v4.8h, v4.8h, #6
srshr v5.8h, v5.8h, #6
srshr v6.8h, v6.8h, #6
srshr v7.8h, v7.8h, #6
uaddw v0.8h, v0.8h, v22.8b
uaddw v1.8h, v1.8h, v23.8b
uaddw v2.8h, v2.8h, v24.8b
uaddw v3.8h, v3.8h, v25.8b
uaddw v4.8h, v4.8h, v26.8b
uaddw v5.8h, v5.8h, v27.8b
uaddw v6.8h, v6.8h, v28.8b
uaddw v7.8h, v7.8h, v29.8b
sqxtun v0.8b, v0.8h
sqxtun v1.8b, v1.8h
sqxtun v2.8b, v2.8h
sqxtun v3.8b, v3.8h
sqxtun v4.8b, v4.8h
sqxtun v5.8b, v5.8h
sqxtun v6.8b, v6.8h
sqxtun v7.8b, v7.8h
st1 {v0.8b}, [x2], x4
st1 {v1.8b}, [x2], x4
st1 {v2.8b}, [x2], x4
st1 {v3.8b}, [x2], x4
st1 {v4.8b}, [x2], x4
st1 {v5.8b}, [x2], x4
st1 {v6.8b}, [x2], x4
st1 {v7.8b}, [x2]
pop_v_regs
ret
|