summaryrefslogtreecommitdiffstats
path: root/kernel/arch/x86/crypto/sha256-ssse3-asm.S
blob: 2cedc44e81216ab89f3fc226dc9b27b2756dfd0b (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
########################################################################
# Implement fast SHA-256 with SSSE3 instructions. (x86_64)
#
# Copyright (C) 2013 Intel Corporation.
#
# Authors:
#     James Guilford <james.guilford@intel.com>
#     Kirk Yap <kirk.s.yap@intel.com>
#     Tim Chen <tim.c.chen@linux.intel.com>
#
# This software is available to you under a choice of one of two
# licenses.  You may choose to be licensed under the terms of the GNU
# General Public License (GPL) Version 2, available from the file
# COPYING in the main directory of this source tree, or the
# OpenIB.org BSD license below:
#
#     Redistribution and use in source and binary forms, with or
#     without modification, are permitted provided that the following
#     conditions are met:
#
#      - Redistributions of source code must retain the above
#        copyright notice, this list of conditions and the following
#        disclaimer.
#
#      - Redistributions in binary form must reproduce the above
#        copyright notice, this list of conditions and the following
#        disclaimer in the documentation and/or other materials
#        provided with the distribution.
#
# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
# EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
# MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
# NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
# BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
# ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
# CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
# SOFTWARE.
#
########################################################################
#
# This code is described in an Intel White-Paper:
# "Fast SHA-256 Implementations on Intel Architecture Processors"
#
# To find it, surf to http://www.intel.com/p/en_US/embedded
# and search for that title.
#
########################################################################

#include <linux/linkage.h>

## assume buffers not aligned
#define    MOVDQ movdqu

################################ Define Macros

# addm [mem], reg
# Add reg to mem using reg-mem add and store
.macro addm p1 p2
        add     \p1, \p2
        mov     \p2, \p1
.endm

################################

# COPY_XMM_AND_BSWAP xmm, [mem], byte_flip_mask
# Load xmm with mem and byte swap each dword
.macro COPY_XMM_AND_BSWAP p1 p2 p3
        MOVDQ \p2, \p1
        pshufb \p3, \p1
.endm

################################

X0 = %xmm4
X1 = %xmm5
X2 = %xmm6
X3 = %xmm7

XTMP0 = %xmm0
XTMP1 = %xmm1
XTMP2 = %xmm2
XTMP3 = %xmm3
XTMP4 = %xmm8
XFER = %xmm9

SHUF_00BA = %xmm10      # shuffle xBxA -> 00BA
SHUF_DC00 = %xmm11      # shuffle xDxC -> DC00
BYTE_FLIP_MASK = %xmm12

NUM_BLKS = %rdx   # 3rd arg
INP = %rsi        # 2nd arg
CTX = %rdi        # 1st arg

SRND = %rsi       # clobbers INP
c = %ecx
d = %r8d
e = %edx
TBL = %rbp
a = %eax
b = %ebx

f = %r9d
g = %r10d
h = %r11d

y0 = %r13d
y1 = %r14d
y2 = %r15d



_INP_END_SIZE = 8
_INP_SIZE = 8
_XFER_SIZE = 16
_XMM_SAVE_SIZE = 0

_INP_END = 0
_INP            = _INP_END  + _INP_END_SIZE
_XFER           = _INP      + _INP_SIZE
_XMM_SAVE       = _XFER     + _XFER_SIZE
STACK_SIZE      = _XMM_SAVE + _XMM_SAVE_SIZE

# rotate_Xs
# Rotate values of symbols X0...X3
.macro rotate_Xs
X_ = X0
X0 = X1
X1 = X2
X2 = X3
X3 = X_
.endm

# ROTATE_ARGS
# Rotate values of symbols a...h
.macro ROTATE_ARGS
TMP_ = h
h = g
g = f
f = e
e = d
d = c
c = b
b = a
a = TMP_
.endm

.macro FOUR_ROUNDS_AND_SCHED
	## compute s0 four at a time and s1 two at a time
	## compute W[-16] + W[-7] 4 at a time
	movdqa  X3, XTMP0
	mov     e, y0			# y0 = e
	ror     $(25-11), y0            # y0 = e >> (25-11)
	mov     a, y1                   # y1 = a
	palignr $4, X2, XTMP0           # XTMP0 = W[-7]
	ror     $(22-13), y1            # y1 = a >> (22-13)
	xor     e, y0                   # y0 = e ^ (e >> (25-11))
	mov     f, y2                   # y2 = f
	ror     $(11-6), y0             # y0 = (e >> (11-6)) ^ (e >> (25-6))
	movdqa  X1, XTMP1
	xor     a, y1                   # y1 = a ^ (a >> (22-13)
	xor     g, y2                   # y2 = f^g
	paddd   X0, XTMP0               # XTMP0 = W[-7] + W[-16]
	xor     e, y0                   # y0 = e ^ (e >> (11-6)) ^ (e >> (25-6))
	and     e, y2                   # y2 = (f^g)&e
	ror     $(13-2), y1             # y1 = (a >> (13-2)) ^ (a >> (22-2))
	## compute s0
	palignr $4, X0, XTMP1           # XTMP1 = W[-15]
	xor     a, y1                   # y1 = a ^ (a >> (13-2)) ^ (a >> (22-2))
	ror     $6, y0                  # y0 = S1 = (e>>6) & (e>>11) ^ (e>>25)
	xor     g, y2                   # y2 = CH = ((f^g)&e)^g
	movdqa  XTMP1, XTMP2            # XTMP2 = W[-15]
	ror     $2, y1                  # y1 = S0 = (a>>2) ^ (a>>13) ^ (a>>22)
	add     y0, y2                  # y2 = S1 + CH
	add     _XFER(%rsp) , y2        # y2 = k + w + S1 + CH
	movdqa  XTMP1, XTMP3            # XTMP3 = W[-15]
	mov     a, y0                   # y0 = a
	add     y2, h                   # h = h + S1 + CH + k + w
	mov     a, y2                   # y2 = a
	pslld   $(32-7), XTMP1          #
	or      c, y0                   # y0 = a|c
	add     h, d                    # d = d + h + S1 + CH + k + w
	and     c, y2                   # y2 = a&c
	psrld   $7, XTMP2               #
	and     b, y0                   # y0 = (a|c)&b
	add     y1, h                   # h = h + S1 + CH + k + w + S0
	por     XTMP2, XTMP1            # XTMP1 = W[-15] ror 7
	or      y2, y0                  # y0 = MAJ = (a|c)&b)|(a&c)
	add     y0, h                   # h = h + S1 + CH + k + w + S0 + MAJ
					#
	ROTATE_ARGS                     #
	movdqa  XTMP3, XTMP2            # XTMP2 = W[-15]
	mov     e, y0                   # y0 = e
	mov     a, y1                   # y1 = a
	movdqa  XTMP3, XTMP4            # XTMP4 = W[-15]
	ror     $(25-11), y0            # y0 = e >> (25-11)
	xor     e, y0                   # y0 = e ^ (e >> (25-11))
	mov     f, y2                   # y2 = f
	ror     $(22-13), y1            # y1 = a >> (22-13)
	pslld   $(32-18), XTMP3         #
	xor     a, y1                   # y1 = a ^ (a >> (22-13)
	ror     $(11-6), y0             # y0 = (e >> (11-6)) ^ (e >> (25-6))
	xor     g, y2                   # y2 = f^g
	psrld   $18, XTMP2              #
	ror     $(13-2), y1             # y1 = (a >> (13-2)) ^ (a >> (22-2))
	xor     e, y0                   # y0 = e ^ (e >> (11-6)) ^ (e >> (25-6))
	and     e, y2                   # y2 = (f^g)&e
	ror     $6, y0                  # y0 = S1 = (e>>6) & (e>>11) ^ (e>>25)
	pxor    XTMP3, XTMP1
	xor     a, y1                   # y1 = a ^ (a >> (13-2)) ^ (a >> (22-2))
	xor     g, y2                   # y2 = CH = ((f^g)&e)^g
	psrld   $3, XTMP4               # XTMP4 = W[-15] >> 3
	add     y0, y2                  # y2 = S1 + CH
	add     (1*4 + _XFER)(%rsp), y2 # y2 = k + w + S1 + CH
	ror     $2, y1                  # y1 = S0 = (a>>2) ^ (a>>13) ^ (a>>22)
	pxor    XTMP2, XTMP1            # XTMP1 = W[-15] ror 7 ^ W[-15] ror 18
	mov     a, y0                   # y0 = a
	add     y2, h                   # h = h + S1 + CH + k + w
	mov     a, y2                   # y2 = a
	pxor    XTMP4, XTMP1            # XTMP1 = s0
	or      c, y0                   # y0 = a|c
	add     h, d                    # d = d + h + S1 + CH + k + w
	and     c, y2                   # y2 = a&c
	## compute low s1
	pshufd  $0b11111010, X3, XTMP2   # XTMP2 = W[-2] {BBAA}
	and     b, y0			# y0 = (a|c)&b
	add     y1, h                   # h = h + S1 + CH + k + w + S0
	paddd   XTMP1, XTMP0            # XTMP0 = W[-16] + W[-7] + s0
	or      y2, y0                  # y0 = MAJ = (a|c)&b)|(a&c)
	add     y0, h                   # h = h + S1 + CH + k + w + S0 + MAJ

	ROTATE_ARGS
	movdqa  XTMP2, XTMP3            # XTMP3 = W[-2] {BBAA}
	mov     e, y0                   # y0 = e
	mov     a, y1                   # y1 = a
	ror     $(25-11), y0            # y0 = e >> (25-11)
	movdqa  XTMP2, XTMP4            # XTMP4 = W[-2] {BBAA}
	xor     e, y0                   # y0 = e ^ (e >> (25-11))
	ror     $(22-13), y1            # y1 = a >> (22-13)
	mov     f, y2                   # y2 = f
	xor     a, y1                   # y1 = a ^ (a >> (22-13)
	ror     $(11-6), y0             # y0 = (e >> (11-6)) ^ (e >> (25-6))
	psrlq   $17, XTMP2              # XTMP2 = W[-2] ror 17 {xBxA}
	xor     g, y2                   # y2 = f^g
	psrlq   $19, XTMP3              # XTMP3 = W[-2] ror 19 {xBxA}
	xor     e, y0                   # y0 = e ^ (e >> (11-6)) ^ (e >> (25-6))
	and     e, y2                   # y2 = (f^g)&e
	psrld   $10, XTMP4              # XTMP4 = W[-2] >> 10 {BBAA}
	ror     $(13-2), y1             # y1 = (a >> (13-2)) ^ (a >> (22-2))
	xor     a, y1                   # y1 = a ^ (a >> (13-2)) ^ (a >> (22-2))
	xor     g, y2                   # y2 = CH = ((f^g)&e)^g
	ror     $6, y0                  # y0 = S1 = (e>>6) & (e>>11) ^ (e>>25)
	pxor    XTMP3, XTMP2
	add     y0, y2                  # y2 = S1 + CH
	ror     $2, y1                  # y1 = S0 = (a>>2) ^ (a>>13) ^ (a>>22)
	add     (2*4 + _XFER)(%rsp), y2 # y2 = k + w + S1 + CH
	pxor    XTMP2, XTMP4            # XTMP4 = s1 {xBxA}
	mov     a, y0                   # y0 = a
	add     y2, h                   # h = h + S1 + CH + k + w
	mov     a, y2                   # y2 = a
	pshufb  SHUF_00BA, XTMP4        # XTMP4 = s1 {00BA}
	or      c, y0                   # y0 = a|c
	add     h, d                    # d = d + h + S1 + CH + k + w
	and     c, y2                   # y2 = a&c
	paddd   XTMP4, XTMP0            # XTMP0 = {..., ..., W[1], W[0]}
	and     b, y0                   # y0 = (a|c)&b
	add     y1, h                   # h = h + S1 + CH + k + w + S0
	## compute high s1
	pshufd  $0b01010000, XTMP0, XTMP2 # XTMP2 = W[-2] {BBAA}
	or      y2, y0                  # y0 = MAJ = (a|c)&b)|(a&c)
	add     y0, h                   # h = h + S1 + CH + k + w + S0 + MAJ
					#
	ROTATE_ARGS                     #
	movdqa  XTMP2, XTMP3            # XTMP3 = W[-2] {DDCC}
	mov     e, y0                   # y0 = e
	ror     $(25-11), y0            # y0 = e >> (25-11)
	mov     a, y1                   # y1 = a
	movdqa  XTMP2, X0               # X0    = W[-2] {DDCC}
	ror     $(22-13), y1            # y1 = a >> (22-13)
	xor     e, y0                   # y0 = e ^ (e >> (25-11))
	mov     f, y2                   # y2 = f
	ror     $(11-6), y0             # y0 = (e >> (11-6)) ^ (e >> (25-6))
	psrlq   $17, XTMP2              # XTMP2 = W[-2] ror 17 {xDxC}
	xor     a, y1                   # y1 = a ^ (a >> (22-13)
	xor     g, y2                   # y2 = f^g
	psrlq   $19, XTMP3              # XTMP3 = W[-2] ror 19 {xDxC}
	xor     e, y0                   # y0 = e ^ (e >> (11-6)) ^ (e >> (25
	and     e, y2                   # y2 = (f^g)&e
	ror     $(13-2), y1             # y1 = (a >> (13-2)) ^ (a >> (22-2))
	psrld   $10, X0                 # X0 = W[-2] >> 10 {DDCC}
	xor     a, y1                   # y1 = a ^ (a >> (13-2)) ^ (a >> (22
	ror     $6, y0                  # y0 = S1 = (e>>6) & (e>>11) ^ (e>>2
	xor     g, y2                   # y2 = CH = ((f^g)&e)^g
	pxor    XTMP3, XTMP2            #
	ror     $2, y1                  # y1 = S0 = (a>>2) ^ (a>>13) ^ (a>>2
	add     y0, y2                  # y2 = S1 + CH
	add     (3*4 + _XFER)(%rsp), y2 # y2 = k + w + S1 + CH
	pxor    XTMP2, X0               # X0 = s1 {xDxC}
	mov     a, y0                   # y0 = a
	add     y2, h                   # h = h + S1 + CH + k + w
	mov     a, y2                   # y2 = a
	pshufb  SHUF_DC00, X0           # X0 = s1 {DC00}
	or      c, y0                   # y0 = a|c
	add     h, d                    # d = d + h + S1 + CH + k + w
	and     c, y2                   # y2 = a&c
	paddd   XTMP0, X0               # X0 = {W[3], W[2], W[1], W[0]}
	and     b, y0                   # y0 = (a|c)&b
	add     y1, h                   # h = h + S1 + CH + k + w + S0
	or      y2, y0                  # y0 = MAJ = (a|c)&b)|(a&c)
	add     y0, h                   # h = h + S1 + CH + k + w + S0 + MAJ

	ROTATE_ARGS
	rotate_Xs
.endm

## input is [rsp + _XFER + %1 * 4]
.macro DO_ROUND round
	mov     e, y0                 # y0 = e
	ror     $(25-11), y0          # y0 = e >> (25-11)
	mov     a, y1                 # y1 = a
	xor     e, y0                 # y0 = e ^ (e >> (25-11))
	ror     $(22-13), y1          # y1 = a >> (22-13)
	mov     f, y2                 # y2 = f
	xor     a, y1                 # y1 = a ^ (a >> (22-13)
	ror     $(11-6), y0           # y0 = (e >> (11-6)) ^ (e >> (25-6))
	xor     g, y2                 # y2 = f^g
	xor     e, y0                 # y0 = e ^ (e >> (11-6)) ^ (e >> (25-6))
	ror     $(13-2), y1           # y1 = (a >> (13-2)) ^ (a >> (22-2))
	and     e, y2                 # y2 = (f^g)&e
	xor     a, y1                 # y1 = a ^ (a >> (13-2)) ^ (a >> (22-2))
	ror     $6, y0                # y0 = S1 = (e>>6) & (e>>11) ^ (e>>25)
	xor     g, y2                 # y2 = CH = ((f^g)&e)^g
	add     y0, y2                # y2 = S1 + CH
	ror     $2, y1                # y1 = S0 = (a>>2) ^ (a>>13) ^ (a>>22)
	offset = \round * 4 + _XFER
	add     offset(%rsp), y2      # y2 = k + w + S1 + CH
	mov     a, y0                 # y0 = a
	add     y2, h                 # h = h + S1 + CH + k + w
	mov     a, y2                 # y2 = a
	or      c, y0                 # y0 = a|c
	add     h, d                  # d = d + h + S1 + CH + k + w
	and     c, y2                 # y2 = a&c
	and     b, y0                 # y0 = (a|c)&b
	add     y1, h                 # h = h + S1 + CH + k + w + S0
	or      y2, y0		      # y0 = MAJ = (a|c)&b)|(a&c)
	add     y0, h		      # h = h + S1 + CH + k + w + S0 + MAJ
	ROTATE_ARGS
.endm

########################################################################
## void sha256_transform_ssse3(void *input_data, UINT32 digest[8], UINT64 num_blks)
## arg 1 : pointer to digest
## arg 2 : pointer to input data
## arg 3 : Num blocks
########################################################################
.text
ENTRY(sha256_transform_ssse3)
.align 32
	pushq   %rbx
	pushq   %rbp
	pushq   %r13
	pushq   %r14
	pushq   %r15
	pushq   %r12

	mov	%rsp, %r12
	subq    $STACK_SIZE, %rsp
	and	$~15, %rsp

	shl     $6, NUM_BLKS		 # convert to bytes
	jz      done_hash
	add     INP, NUM_BLKS
	mov     NUM_BLKS, _INP_END(%rsp) # pointer to end of data

	## load initial digest
	mov     4*0(CTX), a
	mov     4*1(CTX), b
	mov     4*2(CTX), c
	mov     4*3(CTX), d
	mov     4*4(CTX), e
	mov     4*5(CTX), f
	mov     4*6(CTX), g
	mov     4*7(CTX), h

	movdqa  PSHUFFLE_BYTE_FLIP_MASK(%rip), BYTE_FLIP_MASK
	movdqa  _SHUF_00BA(%rip), SHUF_00BA
	movdqa  _SHUF_DC00(%rip), SHUF_DC00

loop0:
	lea     K256(%rip), TBL

	## byte swap first 16 dwords
	COPY_XMM_AND_BSWAP      X0, 0*16(INP), BYTE_FLIP_MASK
	COPY_XMM_AND_BSWAP      X1, 1*16(INP), BYTE_FLIP_MASK
	COPY_XMM_AND_BSWAP      X2, 2*16(INP), BYTE_FLIP_MASK
	COPY_XMM_AND_BSWAP      X3, 3*16(INP), BYTE_FLIP_MASK

	mov     INP, _INP(%rsp)

	## schedule 48 input dwords, by doing 3 rounds of 16 each
	mov     $3, SRND
.align 16
loop1:
	movdqa  (TBL), XFER
	paddd   X0, XFER
	movdqa  XFER, _XFER(%rsp)
	FOUR_ROUNDS_AND_SCHED

	movdqa  1*16(TBL), XFER
	paddd   X0, XFER
	movdqa  XFER, _XFER(%rsp)
	FOUR_ROUNDS_AND_SCHED

	movdqa  2*16(TBL), XFER
	paddd   X0, XFER
	movdqa  XFER, _XFER(%rsp)
	FOUR_ROUNDS_AND_SCHED

	movdqa  3*16(TBL), XFER
	paddd   X0, XFER
	movdqa  XFER, _XFER(%rsp)
	add     $4*16, TBL
	FOUR_ROUNDS_AND_SCHED

	sub     $1, SRND
	jne     loop1

	mov     $2, SRND
loop2:
	paddd   (TBL), X0
	movdqa  X0, _XFER(%rsp)
	DO_ROUND        0
	DO_ROUND        1
	DO_ROUND        2
	DO_ROUND        3
	paddd   1*16(TBL), X1
	movdqa  X1, _XFER(%rsp)
	add     $2*16, TBL
	DO_ROUND        0
	DO_ROUND        1
	DO_ROUND        2
	DO_ROUND        3

	movdqa  X2, X0
	movdqa  X3, X1

	sub     $1, SRND
	jne     loop2

	addm    (4*0)(CTX),a
	addm    (4*1)(CTX),b
	addm    (4*2)(CTX),c
	addm    (4*3)(CTX),d
	addm    (4*4)(CTX),e
	addm    (4*5)(CTX),f
	addm    (4*6)(CTX),g
	addm    (4*7)(CTX),h

	mov     _INP(%rsp), INP
	add     $64, INP
	cmp     _INP_END(%rsp), INP
	jne     loop0

done_hash:

	mov	%r12, %rsp

	popq    %r12
	popq    %r15
	popq    %r14
	popq    %r13
	popq    %rbp
	popq    %rbx

	ret
ENDPROC(sha256_transform_ssse3)

.data
.align 64
K256:
        .long 0x428a2f98,0x71374491,0xb5c0fbcf,0xe9b5dba5
        .long 0x3956c25b,0x59f111f1,0x923f82a4,0xab1c5ed5
        .long 0xd807aa98,0x12835b01,0x243185be,0x550c7dc3
        .long 0x72be5d74,0x80deb1fe,0x9bdc06a7,0xc19bf174
        .long 0xe49b69c1,0xefbe4786,0x0fc19dc6,0x240ca1cc
        .long 0x2de92c6f,0x4a7484aa,0x5cb0a9dc,0x76f988da
        .long 0x983e5152,0xa831c66d,0xb00327c8,0xbf597fc7
        .long 0xc6e00bf3,0xd5a79147,0x06ca6351,0x14292967
        .long 0x27b70a85,0x2e1b2138,0x4d2c6dfc,0x53380d13
        .long 0x650a7354,0x766a0abb,0x81c2c92e,0x92722c85
        .long 0xa2bfe8a1,0xa81a664b,0xc24b8b70,0xc76c51a3
        .long 0xd192e819,0xd6990624,0xf40e3585,0x106aa070
        .long 0x19a4c116,0x1e376c08,0x2748774c,0x34b0bcb5
        .long 0x391c0cb3,0x4ed8aa4a,0x5b9cca4f,0x682e6ff3
        .long 0x748f82ee,0x78a5636f,0x84c87814,0x8cc70208
        .long 0x90befffa,0xa4506ceb,0xbef9a3f7,0xc67178f2

PSHUFFLE_BYTE_FLIP_MASK:
	.octa 0x0c0d0e0f08090a0b0405060700010203

# shuffle xBxA -> 00BA
_SHUF_00BA:
	.octa 0xFFFFFFFFFFFFFFFF0b0a090803020100

# shuffle xDxC -> DC00
_SHUF_DC00:
	.octa 0x0b0a090803020100FFFFFFFFFFFFFFFF