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
|
/*********************************************************************
* Copyright (c) 2022 Loongson Technology Corporation Limited
* Contributed by Gu Xiwei(guxiwei-hf@loongson.cn)
* Shiyou Yin(yinshiyou-hf@loongson.cn)
*
* Permission to use, copy, modify, and/or distribute this software for any
* purpose with or without fee is hereby granted, provided that the above
* copyright notice and this permission notice appear in all copies.
*
* THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
* WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
* MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
* ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
* WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
* ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
* OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
*********************************************************************/
/*
* This file is a LoongArch assembly helper file and available under ISC
* license. It provides a large number of macros and alias to simplify
* writing assembly code, especially for LSX and LASX optimizations.
*
* Any one can modify it or add new features for his/her own purposes.
* Contributing a patch will be appreciated as it might be useful for
* others as well. Send patches to loongson contributor mentioned above.
*
* MAJOR version: Usage changes, incompatible with previous version.
* MINOR version: Add new macros/functions, or bug fixes.
* MICRO version: Comment changes or implementation changes.
*/
#define LML_VERSION_MAJOR 0
#define LML_VERSION_MINOR 4
#define LML_VERSION_MICRO 0
#define DEFAULT_ALIGN 5
/* Set prefix as needed. */
#ifndef PRIVATE_PREFIX
#define PRIVATE_PREFIX dav1d_
#endif
#define PASTE(a,b) a ## b
#define CONCAT(a,b) PASTE(a,b)
#ifdef PREFIX
#define ASM_PREF CONCAT(_,PRIVATE_PREFIX)
#else
#define ASM_PREF PRIVATE_PREFIX
#endif
.macro function name, align=DEFAULT_ALIGN
.macro endfunc
jirl $r0, $r1, 0x0
.size ASM_PREF\name, . - ASM_PREF\name
.purgem endfunc
.endm
.text ;
.align \align ;
.globl ASM_PREF\name ;
.type ASM_PREF\name, @function ;
ASM_PREF\name: ;
.endm
.macro const name, align=DEFAULT_ALIGN
.macro endconst
.size \name, . - \name
.purgem endconst
.endm
.section .rodata
.align \align
\name:
.endm
/*
*============================================================================
* LoongArch register alias
*============================================================================
*/
#define a0 $a0
#define a1 $a1
#define a2 $a2
#define a3 $a3
#define a4 $a4
#define a5 $a5
#define a6 $a6
#define a7 $a7
#define t0 $t0
#define t1 $t1
#define t2 $t2
#define t3 $t3
#define t4 $t4
#define t5 $t5
#define t6 $t6
#define t7 $t7
#define t8 $t8
#define s0 $s0
#define s1 $s1
#define s2 $s2
#define s3 $s3
#define s4 $s4
#define s5 $s5
#define s6 $s6
#define s7 $s7
#define s8 $s8
#define zero $zero
#define sp $sp
#define ra $ra
#define fa0 $fa0
#define fa1 $fa1
#define fa2 $fa2
#define fa3 $fa3
#define fa4 $fa4
#define fa5 $fa5
#define fa6 $fa6
#define fa7 $fa7
#define ft0 $ft0
#define ft1 $ft1
#define ft2 $ft2
#define ft3 $ft3
#define ft4 $ft4
#define ft5 $ft5
#define ft6 $ft6
#define ft7 $ft7
#define ft8 $ft8
#define ft9 $ft9
#define ft10 $ft10
#define ft11 $ft11
#define ft12 $ft12
#define ft13 $ft13
#define ft14 $ft14
#define ft15 $ft15
#define fs0 $fs0
#define fs1 $fs1
#define fs2 $fs2
#define fs3 $fs3
#define fs4 $fs4
#define fs5 $fs5
#define fs6 $fs6
#define fs7 $fs7
#define f0 $f0
#define f1 $f1
#define f2 $f2
#define f3 $f3
#define f4 $f4
#define f5 $f5
#define f6 $f6
#define f7 $f7
#define f8 $f8
#define f9 $f9
#define f10 $f10
#define f11 $f11
#define f12 $f12
#define f13 $f13
#define f14 $f14
#define f15 $f15
#define f16 $f16
#define f17 $f17
#define f18 $f18
#define f19 $f19
#define f20 $f20
#define f21 $f21
#define f22 $f22
#define f23 $f23
#define f24 $f24
#define f25 $f25
#define f26 $f26
#define f27 $f27
#define f28 $f28
#define f29 $f29
#define f30 $f30
#define f31 $f31
#define vr0 $vr0
#define vr1 $vr1
#define vr2 $vr2
#define vr3 $vr3
#define vr4 $vr4
#define vr5 $vr5
#define vr6 $vr6
#define vr7 $vr7
#define vr8 $vr8
#define vr9 $vr9
#define vr10 $vr10
#define vr11 $vr11
#define vr12 $vr12
#define vr13 $vr13
#define vr14 $vr14
#define vr15 $vr15
#define vr16 $vr16
#define vr17 $vr17
#define vr18 $vr18
#define vr19 $vr19
#define vr20 $vr20
#define vr21 $vr21
#define vr22 $vr22
#define vr23 $vr23
#define vr24 $vr24
#define vr25 $vr25
#define vr26 $vr26
#define vr27 $vr27
#define vr28 $vr28
#define vr29 $vr29
#define vr30 $vr30
#define vr31 $vr31
#define xr0 $xr0
#define xr1 $xr1
#define xr2 $xr2
#define xr3 $xr3
#define xr4 $xr4
#define xr5 $xr5
#define xr6 $xr6
#define xr7 $xr7
#define xr8 $xr8
#define xr9 $xr9
#define xr10 $xr10
#define xr11 $xr11
#define xr12 $xr12
#define xr13 $xr13
#define xr14 $xr14
#define xr15 $xr15
#define xr16 $xr16
#define xr17 $xr17
#define xr18 $xr18
#define xr19 $xr19
#define xr20 $xr20
#define xr21 $xr21
#define xr22 $xr22
#define xr23 $xr23
#define xr24 $xr24
#define xr25 $xr25
#define xr26 $xr26
#define xr27 $xr27
#define xr28 $xr28
#define xr29 $xr29
#define xr30 $xr30
#define xr31 $xr31
/*
*============================================================================
* LSX/LASX synthesize instructions
*============================================================================
*/
/*
* Description : Dot product of byte vector elements
* Arguments : Inputs - vj, vk
* Outputs - vd
* Return Type - halfword
*/
.macro vdp2.h.bu vd, vj, vk
vmulwev.h.bu \vd, \vj, \vk
vmaddwod.h.bu \vd, \vj, \vk
.endm
.macro vdp2.h.bu.b vd, vj, vk
vmulwev.h.bu.b \vd, \vj, \vk
vmaddwod.h.bu.b \vd, \vj, \vk
.endm
.macro vdp2.w.h vd, vj, vk
vmulwev.w.h \vd, \vj, \vk
vmaddwod.w.h \vd, \vj, \vk
.endm
.macro xvdp2.h.bu xd, xj, xk
xvmulwev.h.bu \xd, \xj, \xk
xvmaddwod.h.bu \xd, \xj, \xk
.endm
.macro xvdp2.h.bu.b xd, xj, xk
xvmulwev.h.bu.b \xd, \xj, \xk
xvmaddwod.h.bu.b \xd, \xj, \xk
.endm
.macro xvdp2.w.h xd, xj, xk
xvmulwev.w.h \xd, \xj, \xk
xvmaddwod.w.h \xd, \xj, \xk
.endm
/*
* Description : Dot product & addition of halfword vector elements
* Arguments : Inputs - vj, vk
* Outputs - vd
* Return Type - twice size of input
*/
.macro vdp2add.h.bu vd, vj, vk
vmaddwev.h.bu \vd, \vj, \vk
vmaddwod.h.bu \vd, \vj, \vk
.endm
.macro vdp2add.h.bu.b vd, vj, vk
vmaddwev.h.bu.b \vd, \vj, \vk
vmaddwod.h.bu.b \vd, \vj, \vk
.endm
.macro vdp2add.w.h vd, vj, vk
vmaddwev.w.h \vd, \vj, \vk
vmaddwod.w.h \vd, \vj, \vk
.endm
.macro xvdp2add.h.bu.b xd, xj, xk
xvmaddwev.h.bu.b \xd, \xj, \xk
xvmaddwod.h.bu.b \xd, \xj, \xk
.endm
.macro xvdp2add.w.h xd, xj, xk
xvmaddwev.w.h \xd, \xj, \xk
xvmaddwod.w.h \xd, \xj, \xk
.endm
/*
* Description : Range element vj[i] to vk[i] ~ vj[i]
* clip: vj > vk ? vj : vk && vj < va ? vj : va
*/
.macro vclip.h vd, vj, vk, va
vmax.h \vd, \vj, \vk
vmin.h \vd, \vd, \va
.endm
.macro vclip.w vd, vj, vk, va
vmax.w \vd, \vj, \vk
vmin.w \vd, \vd, \va
.endm
.macro xvclip.h xd, xj, xk, xa
xvmax.h \xd, \xj, \xk
xvmin.h \xd, \xd, \xa
.endm
.macro xvclip.w xd, xj, xk, xa
xvmax.w \xd, \xj, \xk
xvmin.w \xd, \xd, \xa
.endm
/*
* Description : Range element vj[i] to 0 ~ 255
* clip255: vj < 255 ? vj : 255 && vj > 0 ? vj : 0
*/
.macro vclip255.h vd, vj
vmaxi.h \vd, \vj, 0
vsat.hu \vd, \vd, 7
.endm
.macro vclip255.w vd, vj
vmaxi.w \vd, \vj, 0
vsat.wu \vd, \vd, 7
.endm
.macro xvclip255.h xd, xj
xvmaxi.h \xd, \xj, 0
xvsat.hu \xd, \xd, 7
.endm
.macro xvclip255.w xd, xj
xvmaxi.w \xd, \xj, 0
xvsat.wu \xd, \xd, 7
.endm
/*
* Description : Store elements of vector
* vd : Data vector to be stroed
* rk : Address of data storage
* ra : Offset of address
* si : Index of data in vd
*/
.macro vstelmx.b vd, rk, ra, si
add.d \rk, \rk, \ra
vstelm.b \vd, \rk, 0, \si
.endm
.macro vstelmx.h vd, rk, ra, si
add.d \rk, \rk, \ra
vstelm.h \vd, \rk, 0, \si
.endm
.macro vstelmx.w vd, rk, ra, si
add.d \rk, \rk, \ra
vstelm.w \vd, \rk, 0, \si
.endm
.macro vstelmx.d vd, rk, ra, si
add.d \rk, \rk, \ra
vstelm.d \vd, \rk, 0, \si
.endm
.macro vmov xd, xj
vor.v \xd, \xj, \xj
.endm
.macro xmov xd, xj
xvor.v \xd, \xj, \xj
.endm
.macro xvstelmx.d xd, rk, ra, si
add.d \rk, \rk, \ra
xvstelm.d \xd, \rk, 0, \si
.endm
/*
*============================================================================
* LSX/LASX custom macros
*============================================================================
*/
/*
* Load 4 float, double, V128, v256 elements with stride.
*/
.macro FLDS_LOADX_4 src, stride, stride2, stride3, out0, out1, out2, out3
fld.s \out0, \src, 0
fldx.s \out1, \src, \stride
fldx.s \out2, \src, \stride2
fldx.s \out3, \src, \stride3
.endm
.macro FLDD_LOADX_4 src, stride, stride2, stride3, out0, out1, out2, out3
fld.d \out0, \src, 0
fldx.d \out1, \src, \stride
fldx.d \out2, \src, \stride2
fldx.d \out3, \src, \stride3
.endm
.macro LSX_LOADX_4 src, stride, stride2, stride3, out0, out1, out2, out3
vld \out0, \src, 0
vldx \out1, \src, \stride
vldx \out2, \src, \stride2
vldx \out3, \src, \stride3
.endm
.macro LASX_LOADX_4 src, stride, stride2, stride3, out0, out1, out2, out3
xvld \out0, \src, 0
xvldx \out1, \src, \stride
xvldx \out2, \src, \stride2
xvldx \out3, \src, \stride3
.endm
/*
* Description : Transpose 4x4 block with half-word elements in vectors
* Arguments : Inputs - in0, in1, in2, in3
* Outputs - out0, out1, out2, out3
*/
.macro LSX_TRANSPOSE4x4_H in0, in1, in2, in3, out0, out1, out2, out3, \
tmp0, tmp1
vilvl.h \tmp0, \in1, \in0
vilvl.h \tmp1, \in3, \in2
vilvl.w \out0, \tmp1, \tmp0
vilvh.w \out2, \tmp1, \tmp0
vilvh.d \out1, \out0, \out0
vilvh.d \out3, \out0, \out2
.endm
/*
* Description : Transpose 4x4 block with word elements in vectors
* Arguments : Inputs - in0, in1, in2, in3
* Outputs - out0, out1, out2, out3
* Details :
* Example :
* 1, 2, 3, 4 1, 5, 9,13
* 5, 6, 7, 8 to 2, 6,10,14
* 9,10,11,12 =====> 3, 7,11,15
* 13,14,15,16 4, 8,12,16
*/
.macro LSX_TRANSPOSE4x4_W in0, in1, in2, in3, out0, out1, out2, out3, \
tmp0, tmp1
vilvl.w \tmp0, \in1, \in0
vilvh.w \out1, \in1, \in0
vilvl.w \tmp1, \in3, \in2
vilvh.w \out3, \in3, \in2
vilvl.d \out0, \tmp1, \tmp0
vilvl.d \out2, \out3, \out1
vilvh.d \out3, \out3, \out1
vilvh.d \out1, \tmp1, \tmp0
.endm
/*
* Description : Transpose 8x8 block with half-word elements in vectors
* Arguments : Inputs - in0, in1, in2, in3, in4, in5, in6, in7
* Outputs - out0, out1, out2, out3, out4, out5, out6, out7
*/
.macro LSX_TRANSPOSE8x8_H in0, in1, in2, in3, in4, in5, in6, in7, out0, out1, \
out2, out3, out4, out5, out6, out7, tmp0, tmp1, tmp2, \
tmp3, tmp4, tmp5, tmp6, tmp7
vilvl.h \tmp0, \in6, \in4
vilvl.h \tmp1, \in7, \in5
vilvl.h \tmp2, \in2, \in0
vilvl.h \tmp3, \in3, \in1
vilvl.h \tmp4, \tmp1, \tmp0
vilvh.h \tmp5, \tmp1, \tmp0
vilvl.h \tmp6, \tmp3, \tmp2
vilvh.h \tmp7, \tmp3, \tmp2
vilvh.h \tmp0, \in6, \in4
vilvh.h \tmp1, \in7, \in5
vilvh.h \tmp2, \in2, \in0
vilvh.h \tmp3, \in3, \in1
vpickev.d \out0, \tmp4, \tmp6
vpickod.d \out1, \tmp4, \tmp6
vpickev.d \out2, \tmp5, \tmp7
vpickod.d \out3, \tmp5, \tmp7
vilvl.h \tmp4, \tmp1, \tmp0
vilvh.h \tmp5, \tmp1, \tmp0
vilvl.h \tmp6, \tmp3, \tmp2
vilvh.h \tmp7, \tmp3, \tmp2
vpickev.d \out4, \tmp4, \tmp6
vpickod.d \out5, \tmp4, \tmp6
vpickev.d \out6, \tmp5, \tmp7
vpickod.d \out7, \tmp5, \tmp7
.endm
/*
* Description : Transpose 16x8 block with byte elements in vectors
* Arguments : Inputs - in0, in1, in2, in3, in4, in5, in6, in7
* Outputs - out0, out1, out2, out3, out4, out5, out6, out7
*/
.macro LASX_TRANSPOSE16X8_B in0, in1, in2, in3, in4, in5, in6, in7, \
in8, in9, in10, in11, in12, in13, in14, in15, \
out0, out1, out2, out3, out4, out5, out6, out7,\
tmp0, tmp1, tmp2, tmp3, tmp4, tmp5, tmp6, tmp7
xvilvl.b \tmp0, \in2, \in0
xvilvl.b \tmp1, \in3, \in1
xvilvl.b \tmp2, \in6, \in4
xvilvl.b \tmp3, \in7, \in5
xvilvl.b \tmp4, \in10, \in8
xvilvl.b \tmp5, \in11, \in9
xvilvl.b \tmp6, \in14, \in12
xvilvl.b \tmp7, \in15, \in13
xvilvl.b \out0, \tmp1, \tmp0
xvilvh.b \out1, \tmp1, \tmp0
xvilvl.b \out2, \tmp3, \tmp2
xvilvh.b \out3, \tmp3, \tmp2
xvilvl.b \out4, \tmp5, \tmp4
xvilvh.b \out5, \tmp5, \tmp4
xvilvl.b \out6, \tmp7, \tmp6
xvilvh.b \out7, \tmp7, \tmp6
xvilvl.w \tmp0, \out2, \out0
xvilvh.w \tmp2, \out2, \out0
xvilvl.w \tmp4, \out3, \out1
xvilvh.w \tmp6, \out3, \out1
xvilvl.w \tmp1, \out6, \out4
xvilvh.w \tmp3, \out6, \out4
xvilvl.w \tmp5, \out7, \out5
xvilvh.w \tmp7, \out7, \out5
xvilvl.d \out0, \tmp1, \tmp0
xvilvh.d \out1, \tmp1, \tmp0
xvilvl.d \out2, \tmp3, \tmp2
xvilvh.d \out3, \tmp3, \tmp2
xvilvl.d \out4, \tmp5, \tmp4
xvilvh.d \out5, \tmp5, \tmp4
xvilvl.d \out6, \tmp7, \tmp6
xvilvh.d \out7, \tmp7, \tmp6
.endm
/*
* Description : Transpose 4x4 block with half-word elements in vectors
* Arguments : Inputs - in0, in1, in2, in3
* Outputs - out0, out1, out2, out3
*/
.macro LASX_TRANSPOSE4x4_H in0, in1, in2, in3, out0, out1, out2, out3, \
tmp0, tmp1
xvilvl.h \tmp0, \in1, \in0
xvilvl.h \tmp1, \in3, \in2
xvilvl.w \out0, \tmp1, \tmp0
xvilvh.w \out2, \tmp1, \tmp0
xvilvh.d \out1, \out0, \out0
xvilvh.d \out3, \out0, \out2
.endm
/*
* Description : Transpose 4x8 block with half-word elements in vectors
* Arguments : Inputs - in0, in1, in2, in3
* Outputs - out0, out1, out2, out3
*/
.macro LASX_TRANSPOSE4x8_H in0, in1, in2, in3, out0, out1, out2, out3, \
tmp0, tmp1
xvilvl.h \tmp0, \in2, \in0
xvilvl.h \tmp1, \in3, \in1
xvilvl.h \out2, \tmp1, \tmp0
xvilvh.h \out3, \tmp1, \tmp0
xvilvl.d \out0, \out2, \out2
xvilvh.d \out1, \out2, \out2
xvilvl.d \out2, \out3, \out3
xvilvh.d \out3, \out3, \out3
.endm
/*
* Description : Transpose 8x8 block with half-word elements in vectors
* Arguments : Inputs - in0, in1, in2, in3, in4, in5, in6, in7
* Outputs - out0, out1, out2, out3, out4, out5, out6, out7
*/
.macro LASX_TRANSPOSE8x8_H in0, in1, in2, in3, in4, in5, in6, in7, \
out0, out1, out2, out3, out4, out5, out6, out7, \
tmp0, tmp1, tmp2, tmp3, tmp4, tmp5, tmp6, tmp7
xvilvl.h \tmp0, \in6, \in4
xvilvl.h \tmp1, \in7, \in5
xvilvl.h \tmp2, \in2, \in0
xvilvl.h \tmp3, \in3, \in1
xvilvl.h \tmp4, \tmp1, \tmp0
xvilvh.h \tmp5, \tmp1, \tmp0
xvilvl.h \tmp6, \tmp3, \tmp2
xvilvh.h \tmp7, \tmp3, \tmp2
xvilvh.h \tmp0, \in6, \in4
xvilvh.h \tmp1, \in7, \in5
xvilvh.h \tmp2, \in2, \in0
xvilvh.h \tmp3, \in3, \in1
xvpickev.d \out0, \tmp4, \tmp6
xvpickod.d \out1, \tmp4, \tmp6
xvpickev.d \out2, \tmp5, \tmp7
xvpickod.d \out3, \tmp5, \tmp7
xvilvl.h \tmp4, \tmp1, \tmp0
xvilvh.h \tmp5, \tmp1, \tmp0
xvilvl.h \tmp6, \tmp3, \tmp2
xvilvh.h \tmp7, \tmp3, \tmp2
xvpickev.d \out4, \tmp4, \tmp6
xvpickod.d \out5, \tmp4, \tmp6
xvpickev.d \out6, \tmp5, \tmp7
xvpickod.d \out7, \tmp5, \tmp7
.endm
/*
* Description : Transpose 2x4x4 block with half-word elements in vectors
* Arguments : Inputs - in0, in1, in2, in3
* Outputs - out0, out1, out2, out3
*/
.macro LASX_TRANSPOSE2x4x4_H in0, in1, in2, in3, out0, out1, out2, out3, \
tmp0, tmp1, tmp2
xvilvh.h \tmp1, \in0, \in1
xvilvl.h \out1, \in0, \in1
xvilvh.h \tmp0, \in2, \in3
xvilvl.h \out3, \in2, \in3
xvilvh.w \tmp2, \out3, \out1
xvilvl.w \out3, \out3, \out1
xvilvl.w \out2, \tmp0, \tmp1
xvilvh.w \tmp1, \tmp0, \tmp1
xvilvh.d \out0, \out2, \out3
xvilvl.d \out2, \out2, \out3
xvilvh.d \out1, \tmp1, \tmp2
xvilvl.d \out3, \tmp1, \tmp2
.endm
/*
* Description : Transpose 4x4 block with word elements in vectors
* Arguments : Inputs - in0, in1, in2, in3
* Outputs - out0, out1, out2, out3
* Details :
* Example :
* 1, 2, 3, 4, 1, 2, 3, 4 1,5, 9,13, 1,5, 9,13
* 5, 6, 7, 8, 5, 6, 7, 8 to 2,6,10,14, 2,6,10,14
* 9,10,11,12, 9,10,11,12 =====> 3,7,11,15, 3,7,11,15
* 13,14,15,16, 13,14,15,16 4,8,12,16, 4,8,12,16
*/
.macro LASX_TRANSPOSE4x4_W in0, in1, in2, in3, out0, out1, out2, out3, \
tmp0, tmp1
xvilvl.w \tmp0, \in1, \in0
xvilvh.w \out1, \in1, \in0
xvilvl.w \tmp1, \in3, \in2
xvilvh.w \out3, \in3, \in2
xvilvl.d \out0, \tmp1, \tmp0
xvilvl.d \out2, \out3, \out1
xvilvh.d \out3, \out3, \out1
xvilvh.d \out1, \tmp1, \tmp0
.endm
/*
* Description : Transpose 8x8 block with word elements in vectors
* Arguments : Inputs - in0, in1, in2, in3, in4, in5, in6, in7
* Outputs - out0, out1, out2, out3, out4, out5, out6,
* _out7
* Example : LASX_TRANSPOSE8x8_W
* in0 : 1,2,3,4,5,6,7,8
* in1 : 2,2,3,4,5,6,7,8
* in2 : 3,2,3,4,5,6,7,8
* in3 : 4,2,3,4,5,6,7,8
* in4 : 5,2,3,4,5,6,7,8
* in5 : 6,2,3,4,5,6,7,8
* in6 : 7,2,3,4,5,6,7,8
* in7 : 8,2,3,4,5,6,7,8
*
* out0 : 1,2,3,4,5,6,7,8
* out1 : 2,2,2,2,2,2,2,2
* out2 : 3,3,3,3,3,3,3,3
* out3 : 4,4,4,4,4,4,4,4
* out4 : 5,5,5,5,5,5,5,5
* out5 : 6,6,6,6,6,6,6,6
* out6 : 7,7,7,7,7,7,7,7
* out7 : 8,8,8,8,8,8,8,8
*/
.macro LASX_TRANSPOSE8x8_W in0, in1, in2, in3, in4, in5, in6, in7,\
out0, out1, out2, out3, out4, out5, out6, out7,\
tmp0, tmp1, tmp2, tmp3
xvilvl.w \tmp0, \in2, \in0
xvilvl.w \tmp1, \in3, \in1
xvilvh.w \tmp2, \in2, \in0
xvilvh.w \tmp3, \in3, \in1
xvilvl.w \out0, \tmp1, \tmp0
xvilvh.w \out1, \tmp1, \tmp0
xvilvl.w \out2, \tmp3, \tmp2
xvilvh.w \out3, \tmp3, \tmp2
xvilvl.w \tmp0, \in6, \in4
xvilvl.w \tmp1, \in7, \in5
xvilvh.w \tmp2, \in6, \in4
xvilvh.w \tmp3, \in7, \in5
xvilvl.w \out4, \tmp1, \tmp0
xvilvh.w \out5, \tmp1, \tmp0
xvilvl.w \out6, \tmp3, \tmp2
xvilvh.w \out7, \tmp3, \tmp2
xmov \tmp0, \out0
xmov \tmp1, \out1
xmov \tmp2, \out2
xmov \tmp3, \out3
xvpermi.q \out0, \out4, 0x02
xvpermi.q \out1, \out5, 0x02
xvpermi.q \out2, \out6, 0x02
xvpermi.q \out3, \out7, 0x02
xvpermi.q \out4, \tmp0, 0x31
xvpermi.q \out5, \tmp1, 0x31
xvpermi.q \out6, \tmp2, 0x31
xvpermi.q \out7, \tmp3, 0x31
.endm
/*
* Description : Transpose 4x4 block with double-word elements in vectors
* Arguments : Inputs - in0, in1, in2, in3
* Outputs - out0, out1, out2, out3
* Example : LASX_TRANSPOSE4x4_D
* in0 : 1,2,3,4
* in1 : 1,2,3,4
* in2 : 1,2,3,4
* in3 : 1,2,3,4
*
* out0 : 1,1,1,1
* out1 : 2,2,2,2
* out2 : 3,3,3,3
* out3 : 4,4,4,4
*/
.macro LASX_TRANSPOSE4x4_D in0, in1, in2, in3, out0, out1, out2, out3, \
tmp0, tmp1
xvilvl.d \tmp0, \in1, \in0
xvilvh.d \out1, \in1, \in0
xvilvh.d \tmp1, \in3, \in2
xvilvl.d \out2, \in3, \in2
xvor.v \out0, \tmp0, \tmp0
xvor.v \out3, \tmp1, \tmp1
xvpermi.q \out0, \out2, 0x02
xvpermi.q \out2, \tmp0, 0x31
xvpermi.q \out3, \out1, 0x31
xvpermi.q \out1, \tmp1, 0x02
.endm
|
