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
|
/* rijndael-aarch64.S - ARMv8/AArch64 assembly implementation of AES cipher
*
* Copyright (C) 2016 Jussi Kivilinna <jussi.kivilinna@iki.fi>
*
* This file is part of Libgcrypt.
*
* Libgcrypt 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.
*
* Libgcrypt 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 this program; if not, see <http://www.gnu.org/licenses/>.
*/
#include "asm-common-aarch64.h"
#if defined(__AARCH64EL__)
#ifdef HAVE_COMPATIBLE_GCC_AARCH64_PLATFORM_AS
.text
/* register macros */
#define CTX x0
#define RDST x1
#define RSRC x2
#define NROUNDS w3
#define RTAB x4
#define RMASK w5
#define RA w8
#define RB w9
#define RC w10
#define RD w11
#define RNA w12
#define RNB w13
#define RNC w14
#define RND w15
#define RT0 w6
#define RT1 w7
#define RT2 w16
#define xRT0 x6
#define xRT1 x7
#define xRT2 x16
#define xw8 x8
#define xw9 x9
#define xw10 x10
#define xw11 x11
#define xw12 x12
#define xw13 x13
#define xw14 x14
#define xw15 x15
/***********************************************************************
* ARMv8/AArch64 assembly implementation of the AES cipher
***********************************************************************/
#define preload_first_key(round, ra) \
ldr ra, [CTX, #(((round) * 16) + 0 * 4)];
#define dummy(round, ra) /* nothing */
#define addroundkey(ra, rb, rc, rd, rna, rnb, rnc, rnd, preload_key) \
ldp rna, rnb, [CTX]; \
ldp rnc, rnd, [CTX, #8]; \
eor ra, ra, rna; \
eor rb, rb, rnb; \
eor rc, rc, rnc; \
preload_key(1, rna); \
eor rd, rd, rnd;
#define do_encround(next_r, ra, rb, rc, rd, rna, rnb, rnc, rnd, preload_key) \
ldr rnb, [CTX, #(((next_r) * 16) + 1 * 4)]; \
\
and RT0, RMASK, ra, lsl#2; \
ldr rnc, [CTX, #(((next_r) * 16) + 2 * 4)]; \
and RT1, RMASK, ra, lsr#(8 - 2); \
ldr rnd, [CTX, #(((next_r) * 16) + 3 * 4)]; \
and RT2, RMASK, ra, lsr#(16 - 2); \
ldr RT0, [RTAB, xRT0]; \
and ra, RMASK, ra, lsr#(24 - 2); \
\
ldr RT1, [RTAB, xRT1]; \
eor rna, rna, RT0; \
ldr RT2, [RTAB, xRT2]; \
and RT0, RMASK, rd, lsl#2; \
ldr ra, [RTAB, x##ra]; \
\
eor rnd, rnd, RT1, ror #24; \
and RT1, RMASK, rd, lsr#(8 - 2); \
eor rnc, rnc, RT2, ror #16; \
and RT2, RMASK, rd, lsr#(16 - 2); \
eor rnb, rnb, ra, ror #8; \
ldr RT0, [RTAB, xRT0]; \
and rd, RMASK, rd, lsr#(24 - 2); \
\
ldr RT1, [RTAB, xRT1]; \
eor rnd, rnd, RT0; \
ldr RT2, [RTAB, xRT2]; \
and RT0, RMASK, rc, lsl#2; \
ldr rd, [RTAB, x##rd]; \
\
eor rnc, rnc, RT1, ror #24; \
and RT1, RMASK, rc, lsr#(8 - 2); \
eor rnb, rnb, RT2, ror #16; \
and RT2, RMASK, rc, lsr#(16 - 2); \
eor rna, rna, rd, ror #8; \
ldr RT0, [RTAB, xRT0]; \
and rc, RMASK, rc, lsr#(24 - 2); \
\
ldr RT1, [RTAB, xRT1]; \
eor rnc, rnc, RT0; \
ldr RT2, [RTAB, xRT2]; \
and RT0, RMASK, rb, lsl#2; \
ldr rc, [RTAB, x##rc]; \
\
eor rnb, rnb, RT1, ror #24; \
and RT1, RMASK, rb, lsr#(8 - 2); \
eor rna, rna, RT2, ror #16; \
and RT2, RMASK, rb, lsr#(16 - 2); \
eor rnd, rnd, rc, ror #8; \
ldr RT0, [RTAB, xRT0]; \
and rb, RMASK, rb, lsr#(24 - 2); \
\
ldr RT1, [RTAB, xRT1]; \
eor rnb, rnb, RT0; \
ldr RT2, [RTAB, xRT2]; \
eor rna, rna, RT1, ror #24; \
ldr rb, [RTAB, x##rb]; \
\
eor rnd, rnd, RT2, ror #16; \
preload_key((next_r) + 1, ra); \
eor rnc, rnc, rb, ror #8;
#define do_lastencround(ra, rb, rc, rd, rna, rnb, rnc, rnd) \
and RT0, RMASK, ra, lsl#2; \
and RT1, RMASK, ra, lsr#(8 - 2); \
and RT2, RMASK, ra, lsr#(16 - 2); \
ldrb rna, [RTAB, xRT0]; \
and ra, RMASK, ra, lsr#(24 - 2); \
ldrb rnd, [RTAB, xRT1]; \
and RT0, RMASK, rd, lsl#2; \
ldrb rnc, [RTAB, xRT2]; \
ror rnd, rnd, #24; \
ldrb rnb, [RTAB, x##ra]; \
and RT1, RMASK, rd, lsr#(8 - 2); \
ror rnc, rnc, #16; \
and RT2, RMASK, rd, lsr#(16 - 2); \
ror rnb, rnb, #8; \
ldrb RT0, [RTAB, xRT0]; \
and rd, RMASK, rd, lsr#(24 - 2); \
ldrb RT1, [RTAB, xRT1]; \
\
orr rnd, rnd, RT0; \
ldrb RT2, [RTAB, xRT2]; \
and RT0, RMASK, rc, lsl#2; \
ldrb rd, [RTAB, x##rd]; \
orr rnc, rnc, RT1, ror #24; \
and RT1, RMASK, rc, lsr#(8 - 2); \
orr rnb, rnb, RT2, ror #16; \
and RT2, RMASK, rc, lsr#(16 - 2); \
orr rna, rna, rd, ror #8; \
ldrb RT0, [RTAB, xRT0]; \
and rc, RMASK, rc, lsr#(24 - 2); \
ldrb RT1, [RTAB, xRT1]; \
\
orr rnc, rnc, RT0; \
ldrb RT2, [RTAB, xRT2]; \
and RT0, RMASK, rb, lsl#2; \
ldrb rc, [RTAB, x##rc]; \
orr rnb, rnb, RT1, ror #24; \
and RT1, RMASK, rb, lsr#(8 - 2); \
orr rna, rna, RT2, ror #16; \
ldrb RT0, [RTAB, xRT0]; \
and RT2, RMASK, rb, lsr#(16 - 2); \
ldrb RT1, [RTAB, xRT1]; \
orr rnd, rnd, rc, ror #8; \
ldrb RT2, [RTAB, xRT2]; \
and rb, RMASK, rb, lsr#(24 - 2); \
ldrb rb, [RTAB, x##rb]; \
\
orr rnb, rnb, RT0; \
orr rna, rna, RT1, ror #24; \
orr rnd, rnd, RT2, ror #16; \
orr rnc, rnc, rb, ror #8;
#define firstencround(round, ra, rb, rc, rd, rna, rnb, rnc, rnd) \
addroundkey(ra, rb, rc, rd, rna, rnb, rnc, rnd, preload_first_key); \
do_encround((round) + 1, ra, rb, rc, rd, rna, rnb, rnc, rnd, preload_first_key);
#define encround(round, ra, rb, rc, rd, rna, rnb, rnc, rnd, preload_key) \
do_encround((round) + 1, ra, rb, rc, rd, rna, rnb, rnc, rnd, preload_key);
#define lastencround(round, ra, rb, rc, rd, rna, rnb, rnc, rnd) \
add CTX, CTX, #(((round) + 1) * 16); \
add RTAB, RTAB, #1; \
do_lastencround(ra, rb, rc, rd, rna, rnb, rnc, rnd); \
addroundkey(rna, rnb, rnc, rnd, ra, rb, rc, rd, dummy);
.globl _gcry_aes_arm_encrypt_block
ELF(.type _gcry_aes_arm_encrypt_block,%function;)
.align 4
_gcry_aes_arm_encrypt_block:
/* input:
* %x0: keysched, CTX
* %x1: dst
* %x2: src
* %w3: number of rounds.. 10, 12 or 14
* %x4: encryption table
*/
CFI_STARTPROC();
/* read input block */
/* aligned load */
ldp RA, RB, [RSRC];
ldp RC, RD, [RSRC, #8];
#ifndef __AARCH64EL__
rev RA, RA;
rev RB, RB;
rev RC, RC;
rev RD, RD;
#endif
mov RMASK, #(0xff<<2);
firstencround(0, RA, RB, RC, RD, RNA, RNB, RNC, RND);
encround(1, RNA, RNB, RNC, RND, RA, RB, RC, RD, preload_first_key);
encround(2, RA, RB, RC, RD, RNA, RNB, RNC, RND, preload_first_key);
encround(3, RNA, RNB, RNC, RND, RA, RB, RC, RD, preload_first_key);
encround(4, RA, RB, RC, RD, RNA, RNB, RNC, RND, preload_first_key);
encround(5, RNA, RNB, RNC, RND, RA, RB, RC, RD, preload_first_key);
encround(6, RA, RB, RC, RD, RNA, RNB, RNC, RND, preload_first_key);
encround(7, RNA, RNB, RNC, RND, RA, RB, RC, RD, preload_first_key);
cmp NROUNDS, #12;
bge .Lenc_not_128;
encround(8, RA, RB, RC, RD, RNA, RNB, RNC, RND, dummy);
lastencround(9, RNA, RNB, RNC, RND, RA, RB, RC, RD);
.Lenc_done:
/* store output block */
/* aligned store */
#ifndef __AARCH64EL__
rev RA, RA;
rev RB, RB;
rev RC, RC;
rev RD, RD;
#endif
/* write output block */
stp RA, RB, [RDST];
stp RC, RD, [RDST, #8];
mov x0, #(0);
ret_spec_stop;
.Lenc_not_128:
beq .Lenc_192
encround(8, RA, RB, RC, RD, RNA, RNB, RNC, RND, preload_first_key);
encround(9, RNA, RNB, RNC, RND, RA, RB, RC, RD, preload_first_key);
encround(10, RA, RB, RC, RD, RNA, RNB, RNC, RND, preload_first_key);
encround(11, RNA, RNB, RNC, RND, RA, RB, RC, RD, preload_first_key);
encround(12, RA, RB, RC, RD, RNA, RNB, RNC, RND, dummy);
lastencround(13, RNA, RNB, RNC, RND, RA, RB, RC, RD);
b .Lenc_done;
.Lenc_192:
encround(8, RA, RB, RC, RD, RNA, RNB, RNC, RND, preload_first_key);
encround(9, RNA, RNB, RNC, RND, RA, RB, RC, RD, preload_first_key);
encround(10, RA, RB, RC, RD, RNA, RNB, RNC, RND, dummy);
lastencround(11, RNA, RNB, RNC, RND, RA, RB, RC, RD);
b .Lenc_done;
CFI_ENDPROC();
ELF(.size _gcry_aes_arm_encrypt_block,.-_gcry_aes_arm_encrypt_block;)
#define addroundkey_dec(round, ra, rb, rc, rd, rna, rnb, rnc, rnd) \
ldr rna, [CTX, #(((round) * 16) + 0 * 4)]; \
ldr rnb, [CTX, #(((round) * 16) + 1 * 4)]; \
eor ra, ra, rna; \
ldr rnc, [CTX, #(((round) * 16) + 2 * 4)]; \
eor rb, rb, rnb; \
ldr rnd, [CTX, #(((round) * 16) + 3 * 4)]; \
eor rc, rc, rnc; \
preload_first_key((round) - 1, rna); \
eor rd, rd, rnd;
#define do_decround(next_r, ra, rb, rc, rd, rna, rnb, rnc, rnd, preload_key) \
ldr rnb, [CTX, #(((next_r) * 16) + 1 * 4)]; \
\
and RT0, RMASK, ra, lsl#2; \
ldr rnc, [CTX, #(((next_r) * 16) + 2 * 4)]; \
and RT1, RMASK, ra, lsr#(8 - 2); \
ldr rnd, [CTX, #(((next_r) * 16) + 3 * 4)]; \
and RT2, RMASK, ra, lsr#(16 - 2); \
ldr RT0, [RTAB, xRT0]; \
and ra, RMASK, ra, lsr#(24 - 2); \
\
ldr RT1, [RTAB, xRT1]; \
eor rna, rna, RT0; \
ldr RT2, [RTAB, xRT2]; \
and RT0, RMASK, rb, lsl#2; \
ldr ra, [RTAB, x##ra]; \
\
eor rnb, rnb, RT1, ror #24; \
and RT1, RMASK, rb, lsr#(8 - 2); \
eor rnc, rnc, RT2, ror #16; \
and RT2, RMASK, rb, lsr#(16 - 2); \
eor rnd, rnd, ra, ror #8; \
ldr RT0, [RTAB, xRT0]; \
and rb, RMASK, rb, lsr#(24 - 2); \
\
ldr RT1, [RTAB, xRT1]; \
eor rnb, rnb, RT0; \
ldr RT2, [RTAB, xRT2]; \
and RT0, RMASK, rc, lsl#2; \
ldr rb, [RTAB, x##rb]; \
\
eor rnc, rnc, RT1, ror #24; \
and RT1, RMASK, rc, lsr#(8 - 2); \
eor rnd, rnd, RT2, ror #16; \
and RT2, RMASK, rc, lsr#(16 - 2); \
eor rna, rna, rb, ror #8; \
ldr RT0, [RTAB, xRT0]; \
and rc, RMASK, rc, lsr#(24 - 2); \
\
ldr RT1, [RTAB, xRT1]; \
eor rnc, rnc, RT0; \
ldr RT2, [RTAB, xRT2]; \
and RT0, RMASK, rd, lsl#2; \
ldr rc, [RTAB, x##rc]; \
\
eor rnd, rnd, RT1, ror #24; \
and RT1, RMASK, rd, lsr#(8 - 2); \
eor rna, rna, RT2, ror #16; \
and RT2, RMASK, rd, lsr#(16 - 2); \
eor rnb, rnb, rc, ror #8; \
ldr RT0, [RTAB, xRT0]; \
and rd, RMASK, rd, lsr#(24 - 2); \
\
ldr RT1, [RTAB, xRT1]; \
eor rnd, rnd, RT0; \
ldr RT2, [RTAB, xRT2]; \
eor rna, rna, RT1, ror #24; \
ldr rd, [RTAB, x##rd]; \
\
eor rnb, rnb, RT2, ror #16; \
preload_key((next_r) - 1, ra); \
eor rnc, rnc, rd, ror #8;
#define do_lastdecround(ra, rb, rc, rd, rna, rnb, rnc, rnd) \
and RT0, RMASK, ra; \
and RT1, RMASK, ra, lsr#8; \
and RT2, RMASK, ra, lsr#16; \
ldrb rna, [RTAB, xRT0]; \
lsr ra, ra, #24; \
ldrb rnb, [RTAB, xRT1]; \
and RT0, RMASK, rb; \
ldrb rnc, [RTAB, xRT2]; \
ror rnb, rnb, #24; \
ldrb rnd, [RTAB, x##ra]; \
and RT1, RMASK, rb, lsr#8; \
ror rnc, rnc, #16; \
and RT2, RMASK, rb, lsr#16; \
ror rnd, rnd, #8; \
ldrb RT0, [RTAB, xRT0]; \
lsr rb, rb, #24; \
ldrb RT1, [RTAB, xRT1]; \
\
orr rnb, rnb, RT0; \
ldrb RT2, [RTAB, xRT2]; \
and RT0, RMASK, rc; \
ldrb rb, [RTAB, x##rb]; \
orr rnc, rnc, RT1, ror #24; \
and RT1, RMASK, rc, lsr#8; \
orr rnd, rnd, RT2, ror #16; \
and RT2, RMASK, rc, lsr#16; \
orr rna, rna, rb, ror #8; \
ldrb RT0, [RTAB, xRT0]; \
lsr rc, rc, #24; \
ldrb RT1, [RTAB, xRT1]; \
\
orr rnc, rnc, RT0; \
ldrb RT2, [RTAB, xRT2]; \
and RT0, RMASK, rd; \
ldrb rc, [RTAB, x##rc]; \
orr rnd, rnd, RT1, ror #24; \
and RT1, RMASK, rd, lsr#8; \
orr rna, rna, RT2, ror #16; \
ldrb RT0, [RTAB, xRT0]; \
and RT2, RMASK, rd, lsr#16; \
ldrb RT1, [RTAB, xRT1]; \
orr rnb, rnb, rc, ror #8; \
ldrb RT2, [RTAB, xRT2]; \
lsr rd, rd, #24; \
ldrb rd, [RTAB, x##rd]; \
\
orr rnd, rnd, RT0; \
orr rna, rna, RT1, ror #24; \
orr rnb, rnb, RT2, ror #16; \
orr rnc, rnc, rd, ror #8;
#define firstdecround(round, ra, rb, rc, rd, rna, rnb, rnc, rnd) \
addroundkey_dec(((round) + 1), ra, rb, rc, rd, rna, rnb, rnc, rnd); \
do_decround(round, ra, rb, rc, rd, rna, rnb, rnc, rnd, preload_first_key);
#define decround(round, ra, rb, rc, rd, rna, rnb, rnc, rnd, preload_key) \
do_decround(round, ra, rb, rc, rd, rna, rnb, rnc, rnd, preload_key);
#define set_last_round_rmask(_, __) \
mov RMASK, #0xff;
#define lastdecround(round, ra, rb, rc, rd, rna, rnb, rnc, rnd) \
add RTAB, RTAB, #(4 * 256); \
do_lastdecround(ra, rb, rc, rd, rna, rnb, rnc, rnd); \
addroundkey(rna, rnb, rnc, rnd, ra, rb, rc, rd, dummy);
.globl _gcry_aes_arm_decrypt_block
ELF(.type _gcry_aes_arm_decrypt_block,%function;)
.align 4
_gcry_aes_arm_decrypt_block:
/* input:
* %x0: keysched, CTX
* %x1: dst
* %x2: src
* %w3: number of rounds.. 10, 12 or 14
* %x4: decryption table
*/
CFI_STARTPROC();
/* read input block */
/* aligned load */
ldp RA, RB, [RSRC];
ldp RC, RD, [RSRC, #8];
#ifndef __AARCH64EL__
rev RA, RA;
rev RB, RB;
rev RC, RC;
rev RD, RD;
#endif
mov RMASK, #(0xff << 2);
cmp NROUNDS, #12;
bge .Ldec_256;
firstdecround(9, RA, RB, RC, RD, RNA, RNB, RNC, RND);
.Ldec_tail:
decround(8, RNA, RNB, RNC, RND, RA, RB, RC, RD, preload_first_key);
decround(7, RA, RB, RC, RD, RNA, RNB, RNC, RND, preload_first_key);
decround(6, RNA, RNB, RNC, RND, RA, RB, RC, RD, preload_first_key);
decround(5, RA, RB, RC, RD, RNA, RNB, RNC, RND, preload_first_key);
decround(4, RNA, RNB, RNC, RND, RA, RB, RC, RD, preload_first_key);
decround(3, RA, RB, RC, RD, RNA, RNB, RNC, RND, preload_first_key);
decround(2, RNA, RNB, RNC, RND, RA, RB, RC, RD, preload_first_key);
decround(1, RA, RB, RC, RD, RNA, RNB, RNC, RND, set_last_round_rmask);
lastdecround(0, RNA, RNB, RNC, RND, RA, RB, RC, RD);
/* store output block */
/* aligned store */
#ifndef __AARCH64EL__
rev RA, RA;
rev RB, RB;
rev RC, RC;
rev RD, RD;
#endif
/* write output block */
stp RA, RB, [RDST];
stp RC, RD, [RDST, #8];
mov x0, #(0);
ret_spec_stop;
.Ldec_256:
beq .Ldec_192;
firstdecround(13, RA, RB, RC, RD, RNA, RNB, RNC, RND);
decround(12, RNA, RNB, RNC, RND, RA, RB, RC, RD, preload_first_key);
decround(11, RA, RB, RC, RD, RNA, RNB, RNC, RND, preload_first_key);
decround(10, RNA, RNB, RNC, RND, RA, RB, RC, RD, preload_first_key);
decround(9, RA, RB, RC, RD, RNA, RNB, RNC, RND, preload_first_key);
b .Ldec_tail;
.Ldec_192:
firstdecround(11, RA, RB, RC, RD, RNA, RNB, RNC, RND);
decround(10, RNA, RNB, RNC, RND, RA, RB, RC, RD, preload_first_key);
decround(9, RA, RB, RC, RD, RNA, RNB, RNC, RND, preload_first_key);
b .Ldec_tail;
CFI_ENDPROC();
ELF(.size _gcry_aes_arm_decrypt_block,.-_gcry_aes_arm_decrypt_block;)
#endif /*HAVE_COMPATIBLE_GCC_AARCH64_PLATFORM_AS*/
#endif /*__AARCH64EL__ */
|
