aboutsummaryrefslogtreecommitdiffstats
path: root/arch/x86/crypto/chacha-avx512vl-x86_64.S
blob: 596f91e3a774ec9dfa0d25968c232632a3d07d0b (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
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
/* SPDX-License-Identifier: GPL-2.0+ */
/*
 * ChaCha 256-bit cipher algorithm, x64 AVX-512VL functions
 *
 * Copyright (C) 2018 Martin Willi
 */

#include <linux/linkage.h>

.section	.rodata.cst32.CTR2BL, "aM", @progbits, 32
.align 32
CTR2BL:	.octa 0x00000000000000000000000000000000
	.octa 0x00000000000000000000000000000001

.section	.rodata.cst32.CTR4BL, "aM", @progbits, 32
.align 32
CTR4BL:	.octa 0x00000000000000000000000000000002
	.octa 0x00000000000000000000000000000003

.section	.rodata.cst32.CTR8BL, "aM", @progbits, 32
.align 32
CTR8BL:	.octa 0x00000003000000020000000100000000
	.octa 0x00000007000000060000000500000004

.text

ENTRY(chacha_2block_xor_avx512vl)
	# %rdi: Input state matrix, s
	# %rsi: up to 2 data blocks output, o
	# %rdx: up to 2 data blocks input, i
	# %rcx: input/output length in bytes
	# %r8d: nrounds

	# This function encrypts two ChaCha blocks by loading the state
	# matrix twice across four AVX registers. It performs matrix operations
	# on four words in each matrix in parallel, but requires shuffling to
	# rearrange the words after each round.

	vzeroupper

	# x0..3[0-2] = s0..3
	vbroadcasti128	0x00(%rdi),%ymm0
	vbroadcasti128	0x10(%rdi),%ymm1
	vbroadcasti128	0x20(%rdi),%ymm2
	vbroadcasti128	0x30(%rdi),%ymm3

	vpaddd		CTR2BL(%rip),%ymm3,%ymm3

	vmovdqa		%ymm0,%ymm8
	vmovdqa		%ymm1,%ymm9
	vmovdqa		%ymm2,%ymm10
	vmovdqa		%ymm3,%ymm11

.Ldoubleround:

	# x0 += x1, x3 = rotl32(x3 ^ x0, 16)
	vpaddd		%ymm1,%ymm0,%ymm0
	vpxord		%ymm0,%ymm3,%ymm3
	vprold		$16,%ymm3,%ymm3

	# x2 += x3, x1 = rotl32(x1 ^ x2, 12)
	vpaddd		%ymm3,%ymm2,%ymm2
	vpxord		%ymm2,%ymm1,%ymm1
	vprold		$12,%ymm1,%ymm1

	# x0 += x1, x3 = rotl32(x3 ^ x0, 8)
	vpaddd		%ymm1,%ymm0,%ymm0
	vpxord		%ymm0,%ymm3,%ymm3
	vprold		$8,%ymm3,%ymm3

	# x2 += x3, x1 = rotl32(x1 ^ x2, 7)
	vpaddd		%ymm3,%ymm2,%ymm2
	vpxord		%ymm2,%ymm1,%ymm1
	vprold		$7,%ymm1,%ymm1

	# x1 = shuffle32(x1, MASK(0, 3, 2, 1))
	vpshufd		$0x39,%ymm1,%ymm1
	# x2 = shuffle32(x2, MASK(1, 0, 3, 2))
	vpshufd		$0x4e,%ymm2,%ymm2
	# x3 = shuffle32(x3, MASK(2, 1, 0, 3))
	vpshufd		$0x93,%ymm3,%ymm3

	# x0 += x1, x3 = rotl32(x3 ^ x0, 16)
	vpaddd		%ymm1,%ymm0,%ymm0
	vpxord		%ymm0,%ymm3,%ymm3
	vprold		$16,%ymm3,%ymm3

	# x2 += x3, x1 = rotl32(x1 ^ x2, 12)
	vpaddd		%ymm3,%ymm2,%ymm2
	vpxord		%ymm2,%ymm1,%ymm1
	vprold		$12,%ymm1,%ymm1

	# x0 += x1, x3 = rotl32(x3 ^ x0, 8)
	vpaddd		%ymm1,%ymm0,%ymm0
	vpxord		%ymm0,%ymm3,%ymm3
	vprold		$8,%ymm3,%ymm3

	# x2 += x3, x1 = rotl32(x1 ^ x2, 7)
	vpaddd		%ymm3,%ymm2,%ymm2
	vpxord		%ymm2,%ymm1,%ymm1
	vprold		$7,%ymm1,%ymm1

	# x1 = shuffle32(x1, MASK(2, 1, 0, 3))
	vpshufd		$0x93,%ymm1,%ymm1
	# x2 = shuffle32(x2, MASK(1, 0, 3, 2))
	vpshufd		$0x4e,%ymm2,%ymm2
	# x3 = shuffle32(x3, MASK(0, 3, 2, 1))
	vpshufd		$0x39,%ymm3,%ymm3

	sub		$2,%r8d
	jnz		.Ldoubleround

	# o0 = i0 ^ (x0 + s0)
	vpaddd		%ymm8,%ymm0,%ymm7
	cmp		$0x10,%rcx
	jl		.Lxorpart2
	vpxord		0x00(%rdx),%xmm7,%xmm6
	vmovdqu		%xmm6,0x00(%rsi)
	vextracti128	$1,%ymm7,%xmm0
	# o1 = i1 ^ (x1 + s1)
	vpaddd		%ymm9,%ymm1,%ymm7
	cmp		$0x20,%rcx
	jl		.Lxorpart2
	vpxord		0x10(%rdx),%xmm7,%xmm6
	vmovdqu		%xmm6,0x10(%rsi)
	vextracti128	$1,%ymm7,%xmm1
	# o2 = i2 ^ (x2 + s2)
	vpaddd		%ymm10,%ymm2,%ymm7
	cmp		$0x30,%rcx
	jl		.Lxorpart2
	vpxord		0x20(%rdx),%xmm7,%xmm6
	vmovdqu		%xmm6,0x20(%rsi)
	vextracti128	$1,%ymm7,%xmm2
	# o3 = i3 ^ (x3 + s3)
	vpaddd		%ymm11,%ymm3,%ymm7
	cmp		$0x40,%rcx
	jl		.Lxorpart2
	vpxord		0x30(%rdx),%xmm7,%xmm6
	vmovdqu		%xmm6,0x30(%rsi)
	vextracti128	$1,%ymm7,%xmm3

	# xor and write second block
	vmovdqa		%xmm0,%xmm7
	cmp		$0x50,%rcx
	jl		.Lxorpart2
	vpxord		0x40(%rdx),%xmm7,%xmm6
	vmovdqu		%xmm6,0x40(%rsi)

	vmovdqa		%xmm1,%xmm7
	cmp		$0x60,%rcx
	jl		.Lxorpart2
	vpxord		0x50(%rdx),%xmm7,%xmm6
	vmovdqu		%xmm6,0x50(%rsi)

	vmovdqa		%xmm2,%xmm7
	cmp		$0x70,%rcx
	jl		.Lxorpart2
	vpxord		0x60(%rdx),%xmm7,%xmm6
	vmovdqu		%xmm6,0x60(%rsi)

	vmovdqa		%xmm3,%xmm7
	cmp		$0x80,%rcx
	jl		.Lxorpart2
	vpxord		0x70(%rdx),%xmm7,%xmm6
	vmovdqu		%xmm6,0x70(%rsi)

.Ldone2:
	vzeroupper
	ret

.Lxorpart2:
	# xor remaining bytes from partial register into output
	mov		%rcx,%rax
	and		$0xf,%rcx
	jz		.Ldone2
	mov		%rax,%r9
	and		$~0xf,%r9

	mov		$1,%rax
	shld		%cl,%rax,%rax
	sub		$1,%rax
	kmovq		%rax,%k1

	vmovdqu8	(%rdx,%r9),%xmm1{%k1}{z}
	vpxord		%xmm7,%xmm1,%xmm1
	vmovdqu8	%xmm1,(%rsi,%r9){%k1}

	jmp		.Ldone2

ENDPROC(chacha_2block_xor_avx512vl)

ENTRY(chacha_4block_xor_avx512vl)
	# %rdi: Input state matrix, s
	# %rsi: up to 4 data blocks output, o
	# %rdx: up to 4 data blocks input, i
	# %rcx: input/output length in bytes
	# %r8d: nrounds

	# This function encrypts four ChaCha blocks by loading the state
	# matrix four times across eight AVX registers. It performs matrix
	# operations on four words in two matrices in parallel, sequentially
	# to the operations on the four words of the other two matrices. The
	# required word shuffling has a rather high latency, we can do the
	# arithmetic on two matrix-pairs without much slowdown.

	vzeroupper

	# x0..3[0-4] = s0..3
	vbroadcasti128	0x00(%rdi),%ymm0
	vbroadcasti128	0x10(%rdi),%ymm1
	vbroadcasti128	0x20(%rdi),%ymm2
	vbroadcasti128	0x30(%rdi),%ymm3

	vmovdqa		%ymm0,%ymm4
	vmovdqa		%ymm1,%ymm5
	vmovdqa		%ymm2,%ymm6
	vmovdqa		%ymm3,%ymm7

	vpaddd		CTR2BL(%rip),%ymm3,%ymm3
	vpaddd		CTR4BL(%rip),%ymm7,%ymm7

	vmovdqa		%ymm0,%ymm11
	vmovdqa		%ymm1,%ymm12
	vmovdqa		%ymm2,%ymm13
	vmovdqa		%ymm3,%ymm14
	vmovdqa		%ymm7,%ymm15

.Ldoubleround4:

	# x0 += x1, x3 = rotl32(x3 ^ x0, 16)
	vpaddd		%ymm1,%ymm0,%ymm0
	vpxord		%ymm0,%ymm3,%ymm3
	vprold		$16,%ymm3,%ymm3

	vpaddd		%ymm5,%ymm4,%ymm4
	vpxord		%ymm4,%ymm7,%ymm7
	vprold		$16,%ymm7,%ymm7

	# x2 += x3, x1 = rotl32(x1 ^ x2, 12)
	vpaddd		%ymm3,%ymm2,%ymm2
	vpxord		%ymm2,%ymm1,%ymm1
	vprold		$12,%ymm1,%ymm1

	vpaddd		%ymm7,%ymm6,%ymm6
	vpxord		%ymm6,%ymm5,%ymm5
	vprold		$12,%ymm5,%ymm5

	# x0 += x1, x3 = rotl32(x3 ^ x0, 8)
	vpaddd		%ymm1,%ymm0,%ymm0
	vpxord		%ymm0,%ymm3,%ymm3
	vprold		$8,%ymm3,%ymm3

	vpaddd		%ymm5,%ymm4,%ymm4
	vpxord		%ymm4,%ymm7,%ymm7
	vprold		$8,%ymm7,%ymm7

	# x2 += x3, x1 = rotl32(x1 ^ x2, 7)
	vpaddd		%ymm3,%ymm2,%ymm2
	vpxord		%ymm2,%ymm1,%ymm1
	vprold		$7,%ymm1,%ymm1

	vpaddd		%ymm7,%ymm6,%ymm6
	vpxord		%ymm6,%ymm5,%ymm5
	vprold		$7,%ymm5,%ymm5

	# x1 = shuffle32(x1, MASK(0, 3, 2, 1))
	vpshufd		$0x39,%ymm1,%ymm1
	vpshufd		$0x39,%ymm5,%ymm5
	# x2 = shuffle32(x2, MASK(1, 0, 3, 2))
	vpshufd		$0x4e,%ymm2,%ymm2
	vpshufd		$0x4e,%ymm6,%ymm6
	# x3 = shuffle32(x3, MASK(2, 1, 0, 3))
	vpshufd		$0x93,%ymm3,%ymm3
	vpshufd		$0x93,%ymm7,%ymm7

	# x0 += x1, x3 = rotl32(x3 ^ x0, 16)
	vpaddd		%ymm1,%ymm0,%ymm0
	vpxord		%ymm0,%ymm3,%ymm3
	vprold		$16,%ymm3,%ymm3

	vpaddd		%ymm5,%ymm4,%ymm4
	vpxord		%ymm4,%ymm7,%ymm7
	vprold		$16,%ymm7,%ymm7

	# x2 += x3, x1 = rotl32(x1 ^ x2, 12)
	vpaddd		%ymm3,%ymm2,%ymm2
	vpxord		%ymm2,%ymm1,%ymm1
	vprold		$12,%ymm1,%ymm1

	vpaddd		%ymm7,%ymm6,%ymm6
	vpxord		%ymm6,%ymm5,%ymm5
	vprold		$12,%ymm5,%ymm5

	# x0 += x1, x3 = rotl32(x3 ^ x0, 8)
	vpaddd		%ymm1,%ymm0,%ymm0
	vpxord		%ymm0,%ymm3,%ymm3
	vprold		$8,%ymm3,%ymm3

	vpaddd		%ymm5,%ymm4,%ymm4
	vpxord		%ymm4,%ymm7,%ymm7
	vprold		$8,%ymm7,%ymm7

	# x2 += x3, x1 = rotl32(x1 ^ x2, 7)
	vpaddd		%ymm3,%ymm2,%ymm2
	vpxord		%ymm2,%ymm1,%ymm1
	vprold		$7,%ymm1,%ymm1

	vpaddd		%ymm7,%ymm6,%ymm6
	vpxord		%ymm6,%ymm5,%ymm5
	vprold		$7,%ymm5,%ymm5

	# x1 = shuffle32(x1, MASK(2, 1, 0, 3))
	vpshufd		$0x93,%ymm1,%ymm1
	vpshufd		$0x93,%ymm5,%ymm5
	# x2 = shuffle32(x2, MASK(1, 0, 3, 2))
	vpshufd		$0x4e,%ymm2,%ymm2
	vpshufd		$0x4e,%ymm6,%ymm6
	# x3 = shuffle32(x3, MASK(0, 3, 2, 1))
	vpshufd		$0x39,%ymm3,%ymm3
	vpshufd		$0x39,%ymm7,%ymm7

	sub		$2,%r8d
	jnz		.Ldoubleround4

	# o0 = i0 ^ (x0 + s0), first block
	vpaddd		%ymm11,%ymm0,%ymm10
	cmp		$0x10,%rcx
	jl		.Lxorpart4
	vpxord		0x00(%rdx),%xmm10,%xmm9
	vmovdqu		%xmm9,0x00(%rsi)
	vextracti128	$1,%ymm10,%xmm0
	# o1 = i1 ^ (x1 + s1), first block
	vpaddd		%ymm12,%ymm1,%ymm10
	cmp		$0x20,%rcx
	jl		.Lxorpart4
	vpxord		0x10(%rdx),%xmm10,%xmm9
	vmovdqu		%xmm9,0x10(%rsi)
	vextracti128	$1,%ymm10,%xmm1
	# o2 = i2 ^ (x2 + s2), first block
	vpaddd		%ymm13,%ymm2,%ymm10
	cmp		$0x30,%rcx
	jl		.Lxorpart4
	vpxord		0x20(%rdx),%xmm10,%xmm9
	vmovdqu		%xmm9,0x20(%rsi)
	vextracti128	$1,%ymm10,%xmm2
	# o3 = i3 ^ (x3 + s3), first block
	vpaddd		%ymm14,%ymm3,%ymm10
	cmp		$0x40,%rcx
	jl		.Lxorpart4
	vpxord		0x30(%rdx),%xmm10,%xmm9
	vmovdqu		%xmm9,0x30(%rsi)
	vextracti128	$1,%ymm10,%xmm3

	# xor and write second block
	vmovdqa		%xmm0,%xmm10
	cmp		$0x50,%rcx
	jl		.Lxorpart4
	vpxord		0x40(%rdx),%xmm10,%xmm9
	vmovdqu		%xmm9,0x40(%rsi)

	vmovdqa		%xmm1,%xmm10
	cmp		$0x60,%rcx
	jl		.Lxorpart4
	vpxord		0x50(%rdx),%xmm10,%xmm9
	vmovdqu		%xmm9,0x50(%rsi)

	vmovdqa		%xmm2,%xmm10
	cmp		$0x70,%rcx
	jl		.Lxorpart4
	vpxord		0x60(%rdx),%xmm10,%xmm9
	vmovdqu		%xmm9,0x60(%rsi)

	vmovdqa		%xmm3,%xmm10
	cmp		$0x80,%rcx
	jl		.Lxorpart4
	vpxord		0x70(%rdx),%xmm10,%xmm9
	vmovdqu		%xmm9,0x70(%rsi)

	# o0 = i0 ^ (x0 + s0), third block
	vpaddd		%ymm11,%ymm4,%ymm10
	cmp		$0x90,%rcx
	jl		.Lxorpart4
	vpxord		0x80(%rdx),%xmm10,%xmm9
	vmovdqu		%xmm9,0x80(%rsi)
	vextracti128	$1,%ymm10,%xmm4
	# o1 = i1 ^ (x1 + s1), third block
	vpaddd		%ymm12,%ymm5,%ymm10
	cmp		$0xa0,%rcx
	jl		.Lxorpart4
	vpxord		0x90(%rdx),%xmm10,%xmm9
	vmovdqu		%xmm9,0x90(%rsi)
	vextracti128	$1,%ymm10,%xmm5
	# o2 = i2 ^ (x2 + s2), third block
	vpaddd		%ymm13,%ymm6,%ymm10
	cmp		$0xb0,%rcx
	jl		.Lxorpart4
	vpxord		0xa0(%rdx),%xmm10,%xmm9
	vmovdqu		%xmm9,0xa0(%rsi)
	vextracti128	$1,%ymm10,%xmm6
	# o3 = i3 ^ (x3 + s3), third block
	vpaddd		%ymm15,%ymm7,%ymm10
	cmp		$0xc0,%rcx
	jl		.Lxorpart4
	vpxord		0xb0(%rdx),%xmm10,%xmm9
	vmovdqu		%xmm9,0xb0(%rsi)
	vextracti128	$1,%ymm10,%xmm7

	# xor and write fourth block
	vmovdqa		%xmm4,%xmm10
	cmp		$0xd0,%rcx
	jl		.Lxorpart4
	vpxord		0xc0(%rdx),%xmm10,%xmm9
	vmovdqu		%xmm9,0xc0(%rsi)

	vmovdqa		%xmm5,%xmm10
	cmp		$0xe0,%rcx
	jl		.Lxorpart4
	vpxord		0xd0(%rdx),%xmm10,%xmm9
	vmovdqu		%xmm9,0xd0(%rsi)

	vmovdqa		%xmm6,%xmm10
	cmp		$0xf0,%rcx
	jl		.Lxorpart4
	vpxord		0xe0(%rdx),%xmm10,%xmm9
	vmovdqu		%xmm9,0xe0(%rsi)

	vmovdqa		%xmm7,%xmm10
	cmp		$0x100,%rcx
	jl		.Lxorpart4
	vpxord		0xf0(%rdx),%xmm10,%xmm9
	vmovdqu		%xmm9,0xf0(%rsi)

.Ldone4:
	vzeroupper
	ret

.Lxorpart4:
	# xor remaining bytes from partial register into output
	mov		%rcx,%rax
	and		$0xf,%rcx
	jz		.Ldone4
	mov		%rax,%r9
	and		$~0xf,%r9

	mov		$1,%rax
	shld		%cl,%rax,%rax
	sub		$1,%rax
	kmovq		%rax,%k1

	vmovdqu8	(%rdx,%r9),%xmm1{%k1}{z}
	vpxord		%xmm10,%xmm1,%xmm1
	vmovdqu8	%xmm1,(%rsi,%r9){%k1}

	jmp		.Ldone4

ENDPROC(chacha_4block_xor_avx512vl)

ENTRY(chacha_8block_xor_avx512vl)
	# %rdi: Input state matrix, s
	# %rsi: up to 8 data blocks output, o
	# %rdx: up to 8 data blocks input, i
	# %rcx: input/output length in bytes
	# %r8d: nrounds

	# This function encrypts eight consecutive ChaCha blocks by loading
	# the state matrix in AVX registers eight times. Compared to AVX2, this
	# mostly benefits from the new rotate instructions in VL and the
	# additional registers.

	vzeroupper

	# x0..15[0-7] = s[0..15]
	vpbroadcastd	0x00(%rdi),%ymm0
	vpbroadcastd	0x04(%rdi),%ymm1
	vpbroadcastd	0x08(%rdi),%ymm2
	vpbroadcastd	0x0c(%rdi),%ymm3
	vpbroadcastd	0x10(%rdi),%ymm4
	vpbroadcastd	0x14(%rdi),%ymm5
	vpbroadcastd	0x18(%rdi),%ymm6
	vpbroadcastd	0x1c(%rdi),%ymm7
	vpbroadcastd	0x20(%rdi),%ymm8
	vpbroadcastd	0x24(%rdi),%ymm9
	vpbroadcastd	0x28(%rdi),%ymm10
	vpbroadcastd	0x2c(%rdi),%ymm11
	vpbroadcastd	0x30(%rdi),%ymm12
	vpbroadcastd	0x34(%rdi),%ymm13
	vpbroadcastd	0x38(%rdi),%ymm14
	vpbroadcastd	0x3c(%rdi),%ymm15

	# x12 += counter values 0-3
	vpaddd		CTR8BL(%rip),%ymm12,%ymm12

	vmovdqa64	%ymm0,%ymm16
	vmovdqa64	%ymm1,%ymm17
	vmovdqa64	%ymm2,%ymm18
	vmovdqa64	%ymm3,%ymm19
	vmovdqa64	%ymm4,%ymm20
	vmovdqa64	%ymm5,%ymm21
	vmovdqa64	%ymm6,%ymm22
	vmovdqa64	%ymm7,%ymm23
	vmovdqa64	%ymm8,%ymm24
	vmovdqa64	%ymm9,%ymm25
	vmovdqa64	%ymm10,%ymm26
	vmovdqa64	%ymm11,%ymm27
	vmovdqa64	%ymm12,%ymm28
	vmovdqa64	%ymm13,%ymm29
	vmovdqa64	%ymm14,%ymm30
	vmovdqa64	%ymm15,%ymm31

.Ldoubleround8:
	# x0 += x4, x12 = rotl32(x12 ^ x0, 16)
	vpaddd		%ymm0,%ymm4,%ymm0
	vpxord		%ymm0,%ymm12,%ymm12
	vprold		$16,%ymm12,%ymm12
	# x1 += x5, x13 = rotl32(x13 ^ x1, 16)
	vpaddd		%ymm1,%ymm5,%ymm1
	vpxord		%ymm1,%ymm13,%ymm13
	vprold		$16,%ymm13,%ymm13
	# x2 += x6, x14 = rotl32(x14 ^ x2, 16)
	vpaddd		%ymm2,%ymm6,%ymm2
	vpxord		%ymm2,%ymm14,%ymm14
	vprold		$16,%ymm14,%ymm14
	# x3 += x7, x15 = rotl32(x15 ^ x3, 16)
	vpaddd		%ymm3,%ymm7,%ymm3
	vpxord		%ymm3,%ymm15,%ymm15
	vprold		$16,%ymm15,%ymm15

	# x8 += x12, x4 = rotl32(x4 ^ x8, 12)
	vpaddd		%ymm12,%ymm8,%ymm8
	vpxord		%ymm8,%ymm4,%ymm4
	vprold		$12,%ymm4,%ymm4
	# x9 += x13, x5 = rotl32(x5 ^ x9, 12)
	vpaddd		%ymm13,%ymm9,%ymm9
	vpxord		%ymm9,%ymm5,%ymm5
	vprold		$12,%ymm5,%ymm5
	# x10 += x14, x6 = rotl32(x6 ^ x10, 12)
	vpaddd		%ymm14,%ymm10,%ymm10
	vpxord		%ymm10,%ymm6,%ymm6
	vprold		$12,%ymm6,%ymm6
	# x11 += x15, x7 = rotl32(x7 ^ x11, 12)
	vpaddd		%ymm15,%ymm11,%ymm11
	vpxord		%ymm11,%ymm7,%ymm7
	vprold		$12,%ymm7,%ymm7

	# x0 += x4, x12 = rotl32(x12 ^ x0, 8)
	vpaddd		%ymm0,%ymm4,%ymm0
	vpxord		%ymm0,%ymm12,%ymm12
	vprold		$8,%ymm12,%ymm12
	# x1 += x5, x13 = rotl32(x13 ^ x1, 8)
	vpaddd		%ymm1,%ymm5,%ymm1
	vpxord		%ymm1,%ymm13,%ymm13
	vprold		$8,%ymm13,%ymm13
	# x2 += x6, x14 = rotl32(x14 ^ x2, 8)
	vpaddd		%ymm2,%ymm6,%ymm2
	vpxord		%ymm2,%ymm14,%ymm14
	vprold		$8,%ymm14,%ymm14
	# x3 += x7, x15 = rotl32(x15 ^ x3, 8)
	vpaddd		%ymm3,%ymm7,%ymm3
	vpxord		%ymm3,%ymm15,%ymm15
	vprold		$8,%ymm15,%ymm15

	# x8 += x12, x4 = rotl32(x4 ^ x8, 7)
	vpaddd		%ymm12,%ymm8,%ymm8
	vpxord		%ymm8,%ymm4,%ymm4
	vprold		$7,%ymm4,%ymm4
	# x9 += x13, x5 = rotl32(x5 ^ x9, 7)
	vpaddd		%ymm13,%ymm9,%ymm9
	vpxord		%ymm9,%ymm5,%ymm5
	vprold		$7,%ymm5,%ymm5
	# x10 += x14, x6 = rotl32(x6 ^ x10, 7)
	vpaddd		%ymm14,%ymm10,%ymm10
	vpxord		%ymm10,%ymm6,%ymm6
	vprold		$7,%ymm6,%ymm6
	# x11 += x15, x7 = rotl32(x7 ^ x11, 7)
	vpaddd		%ymm15,%ymm11,%ymm11
	vpxord		%ymm11,%ymm7,%ymm7
	vprold		$7,%ymm7,%ymm7

	# x0 += x5, x15 = rotl32(x15 ^ x0, 16)
	vpaddd		%ymm0,%ymm5,%ymm0
	vpxord		%ymm0,%ymm15,%ymm15
	vprold		$16,%ymm15,%ymm15
	# x1 += x6, x12 = rotl32(x12 ^ x1, 16)
	vpaddd		%ymm1,%ymm6,%ymm1
	vpxord		%ymm1,%ymm12,%ymm12
	vprold		$16,%ymm12,%ymm12
	# x2 += x7, x13 = rotl32(x13 ^ x2, 16)
	vpaddd		%ymm2,%ymm7,%ymm2
	vpxord		%ymm2,%ymm13,%ymm13
	vprold		$16,%ymm13,%ymm13
	# x3 += x4, x14 = rotl32(x14 ^ x3, 16)
	vpaddd		%ymm3,%ymm4,%ymm3
	vpxord		%ymm3,%ymm14,%ymm14
	vprold		$16,%ymm14,%ymm14

	# x10 += x15, x5 = rotl32(x5 ^ x10, 12)
	vpaddd		%ymm15,%ymm10,%ymm10
	vpxord		%ymm10,%ymm5,%ymm5
	vprold		$12,%ymm5,%ymm5
	# x11 += x12, x6 = rotl32(x6 ^ x11, 12)
	vpaddd		%ymm12,%ymm11,%ymm11
	vpxord		%ymm11,%ymm6,%ymm6
	vprold		$12,%ymm6,%ymm6
	# x8 += x13, x7 = rotl32(x7 ^ x8, 12)
	vpaddd		%ymm13,%ymm8,%ymm8
	vpxord		%ymm8,%ymm7,%ymm7
	vprold		$12,%ymm7,%ymm7
	# x9 += x14, x4 = rotl32(x4 ^ x9, 12)
	vpaddd		%ymm14,%ymm9,%ymm9
	vpxord		%ymm9,%ymm4,%ymm4
	vprold		$12,%ymm4,%ymm4

	# x0 += x5, x15 = rotl32(x15 ^ x0, 8)
	vpaddd		%ymm0,%ymm5,%ymm0
	vpxord		%ymm0,%ymm15,%ymm15
	vprold		$8,%ymm15,%ymm15
	# x1 += x6, x12 = rotl32(x12 ^ x1, 8)
	vpaddd		%ymm1,%ymm6,%ymm1
	vpxord		%ymm1,%ymm12,%ymm12
	vprold		$8,%ymm12,%ymm12
	# x2 += x7, x13 = rotl32(x13 ^ x2, 8)
	vpaddd		%ymm2,%ymm7,%ymm2
	vpxord		%ymm2,%ymm13,%ymm13
	vprold		$8,%ymm13,%ymm13
	# x3 += x4, x14 = rotl32(x14 ^ x3, 8)
	vpaddd		%ymm3,%ymm4,%ymm3
	vpxord		%ymm3,%ymm14,%ymm14
	vprold		$8,%ymm14,%ymm14

	# x10 += x15, x5 = rotl32(x5 ^ x10, 7)
	vpaddd		%ymm15,%ymm10,%ymm10
	vpxord		%ymm10,%ymm5,%ymm5
	vprold		$7,%ymm5,%ymm5
	# x11 += x12, x6 = rotl32(x6 ^ x11, 7)
	vpaddd		%ymm12,%ymm11,%ymm11
	vpxord		%ymm11,%ymm6,%ymm6
	vprold		$7,%ymm6,%ymm6
	# x8 += x13, x7 = rotl32(x7 ^ x8, 7)
	vpaddd		%ymm13,%ymm8,%ymm8
	vpxord		%ymm8,%ymm7,%ymm7
	vprold		$7,%ymm7,%ymm7
	# x9 += x14, x4 = rotl32(x4 ^ x9, 7)
	vpaddd		%ymm14,%ymm9,%ymm9
	vpxord		%ymm9,%ymm4,%ymm4
	vprold		$7,%ymm4,%ymm4

	sub		$2,%r8d
	jnz		.Ldoubleround8

	# x0..15[0-3] += s[0..15]
	vpaddd		%ymm16,%ymm0,%ymm0
	vpaddd		%ymm17,%ymm1,%ymm1
	vpaddd		%ymm18,%ymm2,%ymm2
	vpaddd		%ymm19,%ymm3,%ymm3
	vpaddd		%ymm20,%ymm4,%ymm4
	vpaddd		%ymm21,%ymm5,%ymm5
	vpaddd		%ymm22,%ymm6,%ymm6
	vpaddd		%ymm23,%ymm7,%ymm7
	vpaddd		%ymm24,%ymm8,%ymm8
	vpaddd		%ymm25,%ymm9,%ymm9
	vpaddd		%ymm26,%ymm10,%ymm10
	vpaddd		%ymm27,%ymm11,%ymm11
	vpaddd		%ymm28,%ymm12,%ymm12
	vpaddd		%ymm29,%ymm13,%ymm13
	vpaddd		%ymm30,%ymm14,%ymm14
	vpaddd		%ymm31,%ymm15,%ymm15

	# interleave 32-bit words in state n, n+1
	vpunpckldq	%ymm1,%ymm0,%ymm16
	vpunpckhdq	%ymm1,%ymm0,%ymm17
	vpunpckldq	%ymm3,%ymm2,%ymm18
	vpunpckhdq	%ymm3,%ymm2,%ymm19
	vpunpckldq	%ymm5,%ymm4,%ymm20
	vpunpckhdq	%ymm5,%ymm4,%ymm21
	vpunpckldq	%ymm7,%ymm6,%ymm22
	vpunpckhdq	%ymm7,%ymm6,%ymm23
	vpunpckldq	%ymm9,%ymm8,%ymm24
	vpunpckhdq	%ymm9,%ymm8,%ymm25
	vpunpckldq	%ymm11,%ymm10,%ymm26
	vpunpckhdq	%ymm11,%ymm10,%ymm27
	vpunpckldq	%ymm13,%ymm12,%ymm28
	vpunpckhdq	%ymm13,%ymm12,%ymm29
	vpunpckldq	%ymm15,%ymm14,%ymm30
	vpunpckhdq	%ymm15,%ymm14,%ymm31

	# interleave 64-bit words in state n, n+2
	vpunpcklqdq	%ymm18,%ymm16,%ymm0
	vpunpcklqdq	%ymm19,%ymm17,%ymm1
	vpunpckhqdq	%ymm18,%ymm16,%ymm2
	vpunpckhqdq	%ymm19,%ymm17,%ymm3
	vpunpcklqdq	%ymm22,%ymm20,%ymm4
	vpunpcklqdq	%ymm23,%ymm21,%ymm5
	vpunpckhqdq	%ymm22,%ymm20,%ymm6
	vpunpckhqdq	%ymm23,%ymm21,%ymm7
	vpunpcklqdq	%ymm26,%ymm24,%ymm8
	vpunpcklqdq	%ymm27,%ymm25,%ymm9
	vpunpckhqdq	%ymm26,%ymm24,%ymm10
	vpunpckhqdq	%ymm27,%ymm25,%ymm11
	vpunpcklqdq	%ymm30,%ymm28,%ymm12
	vpunpcklqdq	%ymm31,%ymm29,%ymm13
	vpunpckhqdq	%ymm30,%ymm28,%ymm14
	vpunpckhqdq	%ymm31,%ymm29,%ymm15

	# interleave 128-bit words in state n, n+4
	# xor/write first four blocks
	vmovdqa64	%ymm0,%ymm16
	vperm2i128	$0x20,%ymm4,%ymm0,%ymm0
	cmp		$0x0020,%rcx
	jl		.Lxorpart8
	vpxord		0x0000(%rdx),%ymm0,%ymm0
	vmovdqu64	%ymm0,0x0000(%rsi)
	vmovdqa64	%ymm16,%ymm0
	vperm2i128	$0x31,%ymm4,%ymm0,%ymm4

	vperm2i128	$0x20,%ymm12,%ymm8,%ymm0
	cmp		$0x0040,%rcx
	jl		.Lxorpart8
	vpxord		0x0020(%rdx),%ymm0,%ymm0
	vmovdqu64	%ymm0,0x0020(%rsi)
	vperm2i128	$0x31,%ymm12,%ymm8,%ymm12

	vperm2i128	$0x20,%ymm6,%ymm2,%ymm0
	cmp		$0x0060,%rcx
	jl		.Lxorpart8
	vpxord		0x0040(%rdx),%ymm0,%ymm0
	vmovdqu64	%ymm0,0x0040(%rsi)
	vperm2i128	$0x31,%ymm6,%ymm2,%ymm6

	vperm2i128	$0x20,%ymm14,%ymm10,%ymm0
	cmp		$0x0080,%rcx
	jl		.Lxorpart8
	vpxord		0x0060(%rdx),%ymm0,%ymm0
	vmovdqu64	%ymm0,0x0060(%rsi)
	vperm2i128	$0x31,%ymm14,%ymm10,%ymm14

	vperm2i128	$0x20,%ymm5,%ymm1,%ymm0
	cmp		$0x00a0,%rcx
	jl		.Lxorpart8
	vpxord		0x0080(%rdx),%ymm0,%ymm0
	vmovdqu64	%ymm0,0x0080(%rsi)
	vperm2i128	$0x31,%ymm5,%ymm1,%ymm5

	vperm2i128	$0x20,%ymm13,%ymm9,%ymm0
	cmp		$0x00c0,%rcx
	jl		.Lxorpart8
	vpxord		0x00a0(%rdx),%ymm0,%ymm0
	vmovdqu64	%ymm0,0x00a0(%rsi)
	vperm2i128	$0x31,%ymm13,%ymm9,%ymm13

	vperm2i128	$0x20,%ymm7,%ymm3,%ymm0
	cmp		$0x00e0,%rcx
	jl		.Lxorpart8
	vpxord		0x00c0(%rdx),%ymm0,%ymm0
	vmovdqu64	%ymm0,0x00c0(%rsi)
	vperm2i128	$0x31,%ymm7,%ymm3,%ymm7

	vperm2i128	$0x20,%ymm15,%ymm11,%ymm0
	cmp		$0x0100,%rcx
	jl		.Lxorpart8
	vpxord		0x00e0(%rdx),%ymm0,%ymm0
	vmovdqu64	%ymm0,0x00e0(%rsi)
	vperm2i128	$0x31,%ymm15,%ymm11,%ymm15

	# xor remaining blocks, write to output
	vmovdqa64	%ymm4,%ymm0
	cmp		$0x0120,%rcx
	jl		.Lxorpart8
	vpxord		0x0100(%rdx),%ymm0,%ymm0
	vmovdqu64	%ymm0,0x0100(%rsi)

	vmovdqa64	%ymm12,%ymm0
	cmp		$0x0140,%rcx
	jl		.Lxorpart8
	vpxord		0x0120(%rdx),%ymm0,%ymm0
	vmovdqu64	%ymm0,0x0120(%rsi)

	vmovdqa64	%ymm6,%ymm0
	cmp		$0x0160,%rcx
	jl		.Lxorpart8
	vpxord		0x0140(%rdx),%ymm0,%ymm0
	vmovdqu64	%ymm0,0x0140(%rsi)

	vmovdqa64	%ymm14,%ymm0
	cmp		$0x0180,%rcx
	jl		.Lxorpart8
	vpxord		0x0160(%rdx),%ymm0,%ymm0
	vmovdqu64	%ymm0,0x0160(%rsi)

	vmovdqa64	%ymm5,%ymm0
	cmp		$0x01a0,%rcx
	jl		.Lxorpart8
	vpxord		0x0180(%rdx),%ymm0,%ymm0
	vmovdqu64	%ymm0,0x0180(%rsi)

	vmovdqa64	%ymm13,%ymm0
	cmp		$0x01c0,%rcx
	jl		.Lxorpart8
	vpxord		0x01a0(%rdx),%ymm0,%ymm0
	vmovdqu64	%ymm0,0x01a0(%rsi)

	vmovdqa64	%ymm7,%ymm0
	cmp		$0x01e0,%rcx
	jl		.Lxorpart8
	vpxord		0x01c0(%rdx),%ymm0,%ymm0
	vmovdqu64	%ymm0,0x01c0(%rsi)

	vmovdqa64	%ymm15,%ymm0
	cmp		$0x0200,%rcx
	jl		.Lxorpart8
	vpxord		0x01e0(%rdx),%ymm0,%ymm0
	vmovdqu64	%ymm0,0x01e0(%rsi)

.Ldone8:
	vzeroupper
	ret

.Lxorpart8:
	# xor remaining bytes from partial register into output
	mov		%rcx,%rax
	and		$0x1f,%rcx
	jz		.Ldone8
	mov		%rax,%r9
	and		$~0x1f,%r9

	mov		$1,%rax
	shld		%cl,%rax,%rax
	sub		$1,%rax
	kmovq		%rax,%k1

	vmovdqu8	(%rdx,%r9),%ymm1{%k1}{z}
	vpxord		%ymm0,%ymm1,%ymm1
	vmovdqu8	%ymm1,(%rsi,%r9){%k1}

	jmp		.Ldone8

ENDPROC(chacha_8block_xor_avx512vl)