aboutsummaryrefslogtreecommitdiff
path: root/sys/crypto/armv8/armv8_crypto_wrap.c
blob: eb4a431d33e9a0e95fcfc93ed193974bb7744217 (plain) (blame)
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
/*-
 * Copyright (c) 2016 The FreeBSD Foundation
 * Copyright (c) 2020 Ampere Computing
 * All rights reserved.
 *
 * This software was developed by Andrew Turner under
 * sponsorship from the FreeBSD Foundation.
 *
 * Redistribution and use in source and binary forms, with or without
 * modification, are permitted provided that the following conditions
 * are met:
 * 1. Redistributions of source code must retain the above copyright
 *    notice, this list of conditions and the following disclaimer.
 * 2. 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.
 *
 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
 * ARE DISCLAIMED.  IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
 * SUCH DAMAGE.
 */

/*
 * This code is built with floating-point enabled. Make sure to have entered
 * into floating-point context before calling any of these functions.
 */

#include <sys/cdefs.h>
__FBSDID("$FreeBSD$");

#include <sys/param.h>
#include <sys/systm.h>
#include <sys/malloc.h>
#include <sys/queue.h>

#include <opencrypto/cryptodev.h>
#include <opencrypto/gmac.h>
#include <crypto/rijndael/rijndael.h>
#include <crypto/armv8/armv8_crypto.h>

#include <arm_neon.h>

static uint8x16_t
armv8_aes_enc(int rounds, const uint8x16_t *keysched, const uint8x16_t from)
{
	uint8x16_t tmp;
	int i;

	tmp = from;
	for (i = 0; i < rounds - 1; i += 2) {
		tmp = vaeseq_u8(tmp, keysched[i]);
		tmp = vaesmcq_u8(tmp);
		tmp = vaeseq_u8(tmp, keysched[i + 1]);
		tmp = vaesmcq_u8(tmp);
	}

	tmp = vaeseq_u8(tmp, keysched[rounds - 1]);
	tmp = vaesmcq_u8(tmp);
	tmp = vaeseq_u8(tmp, keysched[rounds]);
	tmp = veorq_u8(tmp, keysched[rounds + 1]);

	return (tmp);
}

static uint8x16_t
armv8_aes_dec(int rounds, const uint8x16_t *keysched, const uint8x16_t from)
{
	uint8x16_t tmp;
	int i;

	tmp = from;
	for (i = 0; i < rounds - 1; i += 2) {
		tmp = vaesdq_u8(tmp, keysched[i]);
		tmp = vaesimcq_u8(tmp);
		tmp = vaesdq_u8(tmp, keysched[i+1]);
		tmp = vaesimcq_u8(tmp);
	}

	tmp = vaesdq_u8(tmp, keysched[rounds - 1]);
	tmp = vaesimcq_u8(tmp);
	tmp = vaesdq_u8(tmp, keysched[rounds]);
	tmp = veorq_u8(tmp, keysched[rounds + 1]);

	return (tmp);
}

void
armv8_aes_encrypt_cbc(const AES_key_t *key, size_t len,
    const uint8_t *from, uint8_t *to, const uint8_t iv[static AES_BLOCK_LEN])
{
	uint8x16_t tot, ivreg, tmp;
	size_t i;

	len /= AES_BLOCK_LEN;
	ivreg = vld1q_u8(iv);
	for (i = 0; i < len; i++) {
		tmp = vld1q_u8(from);
		tot = armv8_aes_enc(key->aes_rounds - 1,
		    (const void*)key->aes_key, veorq_u8(tmp, ivreg));
		ivreg = tot;
		vst1q_u8(to, tot);
		from += AES_BLOCK_LEN;
		to += AES_BLOCK_LEN;
	}
}

void
armv8_aes_decrypt_cbc(const AES_key_t *key, size_t len,
    uint8_t *buf, const uint8_t iv[static AES_BLOCK_LEN])
{
	uint8x16_t ivreg, nextiv, tmp;
	size_t i;

	len /= AES_BLOCK_LEN;
	ivreg = vld1q_u8(iv);
	for (i = 0; i < len; i++) {
		nextiv = vld1q_u8(buf);
		tmp = armv8_aes_dec(key->aes_rounds - 1,
		    (const void*)key->aes_key, nextiv);
		vst1q_u8(buf, veorq_u8(tmp, ivreg));
		ivreg = nextiv;
		buf += AES_BLOCK_LEN;
	}
}

#define	AES_XTS_BLOCKSIZE	16
#define	AES_XTS_IVSIZE		8
#define	AES_XTS_ALPHA		0x87	/* GF(2^128) generator polynomial */

static inline int32x4_t
xts_crank_lfsr(int32x4_t inp)
{
	const int32x4_t alphamask = {AES_XTS_ALPHA, 1, 1, 1};
	int32x4_t xtweak, ret;

	/* set up xor mask */
	xtweak = vextq_s32(inp, inp, 3);
	xtweak = vshrq_n_s32(xtweak, 31);
	xtweak &= alphamask;

	/* next term */
	ret = vshlq_n_s32(inp, 1);
	ret ^= xtweak;

	return ret;
}

static void
armv8_aes_crypt_xts_block(int rounds, const uint8x16_t *key_schedule,
    uint8x16_t *tweak, const uint8_t *from, uint8_t *to, int do_encrypt)
{
	uint8x16_t block;

	block = vld1q_u8(from) ^ *tweak;

	if (do_encrypt)
		block = armv8_aes_enc(rounds - 1, key_schedule, block);
	else
		block = armv8_aes_dec(rounds - 1, key_schedule, block);

	vst1q_u8(to, block ^ *tweak);

	*tweak = vreinterpretq_u8_s32(xts_crank_lfsr(vreinterpretq_s32_u8(*tweak)));
}

static void
armv8_aes_crypt_xts(int rounds, const uint8x16_t *data_schedule,
    const uint8x16_t *tweak_schedule, size_t len, const uint8_t *from,
    uint8_t *to, const uint8_t iv[static AES_BLOCK_LEN], int do_encrypt)
{
	uint8x16_t tweakreg;
	uint8_t tweak[AES_XTS_BLOCKSIZE] __aligned(16);
	size_t i, cnt;

	/*
	 * Prepare tweak as E_k2(IV). IV is specified as LE representation
	 * of a 64-bit block number which we allow to be passed in directly.
	 */
#if BYTE_ORDER == LITTLE_ENDIAN
	bcopy(iv, tweak, AES_XTS_IVSIZE);
	/* Last 64 bits of IV are always zero. */
	bzero(tweak + AES_XTS_IVSIZE, AES_XTS_IVSIZE);
#else
#error Only LITTLE_ENDIAN architectures are supported.
#endif
	tweakreg = vld1q_u8(tweak);
	tweakreg = armv8_aes_enc(rounds - 1, tweak_schedule, tweakreg);

	cnt = len / AES_XTS_BLOCKSIZE;
	for (i = 0; i < cnt; i++) {
		armv8_aes_crypt_xts_block(rounds, data_schedule, &tweakreg,
		    from, to, do_encrypt);
		from += AES_XTS_BLOCKSIZE;
		to += AES_XTS_BLOCKSIZE;
	}
}

void
armv8_aes_encrypt_xts(AES_key_t *data_schedule,
    const void *tweak_schedule, size_t len, const uint8_t *from, uint8_t *to,
    const uint8_t iv[static AES_BLOCK_LEN])
{

	armv8_aes_crypt_xts(data_schedule->aes_rounds,
	    (const void *)&data_schedule->aes_key, tweak_schedule, len, from,
	    to, iv, 1);
}

void
armv8_aes_decrypt_xts(AES_key_t *data_schedule,
    const void *tweak_schedule, size_t len, const uint8_t *from, uint8_t *to,
    const uint8_t iv[static AES_BLOCK_LEN])
{

	armv8_aes_crypt_xts(data_schedule->aes_rounds,
	    (const void *)&data_schedule->aes_key, tweak_schedule, len, from,
	    to,iv, 0);

}

#define	AES_INC_COUNTER(counter)				\
	do {							\
		for (int pos = AES_BLOCK_LEN - 1;		\
		     pos >= 0; pos--)				\
			if (++(counter)[pos])			\
				break;				\
	} while (0)

struct armv8_gcm_state {
	__uint128_val_t EK0;
	__uint128_val_t EKi;
	__uint128_val_t Xi;
	__uint128_val_t lenblock;
	uint8_t aes_counter[AES_BLOCK_LEN];
};

void
armv8_aes_encrypt_gcm(AES_key_t *aes_key, size_t len,
    const uint8_t *from, uint8_t *to,
    size_t authdatalen, const uint8_t *authdata,
    uint8_t tag[static GMAC_DIGEST_LEN],
    const uint8_t iv[static AES_GCM_IV_LEN],
    const __uint128_val_t *Htable)
{
	struct armv8_gcm_state s;
	const uint64_t *from64;
	uint64_t *to64;
	uint8_t block[AES_BLOCK_LEN];
	size_t i, trailer;

	bzero(&s.aes_counter, AES_BLOCK_LEN);
	memcpy(s.aes_counter, iv, AES_GCM_IV_LEN);

	/* Setup the counter */
	s.aes_counter[AES_BLOCK_LEN - 1] = 1;

	/* EK0 for a final GMAC round */
	aes_v8_encrypt(s.aes_counter, s.EK0.c, aes_key);

	/* GCM starts with 2 as counter, 1 is used for final xor of tag. */
	s.aes_counter[AES_BLOCK_LEN - 1] = 2;

	memset(s.Xi.c, 0, sizeof(s.Xi.c));
	trailer = authdatalen % AES_BLOCK_LEN;
	if (authdatalen - trailer > 0) {
		gcm_ghash_v8(s.Xi.u, Htable, authdata, authdatalen - trailer);
		authdata += authdatalen - trailer;
	}
	if (trailer > 0 || authdatalen == 0) {
		memset(block, 0, sizeof(block));
		memcpy(block, authdata, trailer);
		gcm_ghash_v8(s.Xi.u, Htable, block, AES_BLOCK_LEN);
	}

	from64 = (const uint64_t*)from;
	to64 = (uint64_t*)to;
	trailer = len % AES_BLOCK_LEN;

	for (i = 0; i < (len - trailer); i += AES_BLOCK_LEN) {
		aes_v8_encrypt(s.aes_counter, s.EKi.c, aes_key);
		AES_INC_COUNTER(s.aes_counter);
		to64[0] = from64[0] ^ s.EKi.u[0];
		to64[1] = from64[1] ^ s.EKi.u[1];
		gcm_ghash_v8(s.Xi.u, Htable, (uint8_t*)to64, AES_BLOCK_LEN);

		to64 += 2;
		from64 += 2;
	}

	to += (len - trailer);
	from += (len - trailer);

	if (trailer) {
		aes_v8_encrypt(s.aes_counter, s.EKi.c, aes_key);
		AES_INC_COUNTER(s.aes_counter);
		memset(block, 0, sizeof(block));
		for (i = 0; i < trailer; i++) {
			block[i] = to[i] = from[i] ^ s.EKi.c[i];
		}

		gcm_ghash_v8(s.Xi.u, Htable, block, AES_BLOCK_LEN);
	}

	/* Lengths block */
	s.lenblock.u[0] = s.lenblock.u[1] = 0;
	s.lenblock.d[1] = htobe32(authdatalen * 8);
	s.lenblock.d[3] = htobe32(len * 8);
	gcm_ghash_v8(s.Xi.u, Htable, s.lenblock.c, AES_BLOCK_LEN);

	s.Xi.u[0] ^= s.EK0.u[0];
	s.Xi.u[1] ^= s.EK0.u[1];
	memcpy(tag, s.Xi.c, GMAC_DIGEST_LEN);

	explicit_bzero(&s, sizeof(s));
}

int
armv8_aes_decrypt_gcm(AES_key_t *aes_key, size_t len,
    const uint8_t *from, uint8_t *to,
    size_t authdatalen, const uint8_t *authdata,
    const uint8_t tag[static GMAC_DIGEST_LEN],
    const uint8_t iv[static AES_GCM_IV_LEN],
    const __uint128_val_t *Htable)
{
	struct armv8_gcm_state s;
	const uint64_t *from64;
	uint64_t *to64;
	uint8_t block[AES_BLOCK_LEN];
	size_t i, trailer;
	int error;

	error = 0;
	bzero(&s.aes_counter, AES_BLOCK_LEN);
	memcpy(s.aes_counter, iv, AES_GCM_IV_LEN);

	/* Setup the counter */
	s.aes_counter[AES_BLOCK_LEN - 1] = 1;

	/* EK0 for a final GMAC round */
	aes_v8_encrypt(s.aes_counter, s.EK0.c, aes_key);

	memset(s.Xi.c, 0, sizeof(s.Xi.c));
	trailer = authdatalen % AES_BLOCK_LEN;
	if (authdatalen - trailer > 0) {
		gcm_ghash_v8(s.Xi.u, Htable, authdata, authdatalen - trailer);
		authdata += authdatalen - trailer;
	}
	if (trailer > 0 || authdatalen == 0) {
		memset(block, 0, sizeof(block));
		memcpy(block, authdata, trailer);
		gcm_ghash_v8(s.Xi.u, Htable, block, AES_BLOCK_LEN);
	}

	trailer = len % AES_BLOCK_LEN;
	if (len - trailer > 0)
		gcm_ghash_v8(s.Xi.u, Htable, from, len - trailer);
	if (trailer > 0) {
		memset(block, 0, sizeof(block));
		memcpy(block, from + len - trailer, trailer);
		gcm_ghash_v8(s.Xi.u, Htable, block, AES_BLOCK_LEN);
	}

	/* Lengths block */
	s.lenblock.u[0] = s.lenblock.u[1] = 0;
	s.lenblock.d[1] = htobe32(authdatalen * 8);
	s.lenblock.d[3] = htobe32(len * 8);
	gcm_ghash_v8(s.Xi.u, Htable, s.lenblock.c, AES_BLOCK_LEN);

	s.Xi.u[0] ^= s.EK0.u[0];
	s.Xi.u[1] ^= s.EK0.u[1];
	if (timingsafe_bcmp(tag, s.Xi.c, GMAC_DIGEST_LEN) != 0) {
		error = EBADMSG;
		goto out;
	}

	/* GCM starts with 2 as counter, 1 is used for final xor of tag. */
	s.aes_counter[AES_BLOCK_LEN - 1] = 2;

	from64 = (const uint64_t*)from;
	to64 = (uint64_t*)to;

	for (i = 0; i < (len - trailer); i += AES_BLOCK_LEN) {
		aes_v8_encrypt(s.aes_counter, s.EKi.c, aes_key);
		AES_INC_COUNTER(s.aes_counter);
		to64[0] = from64[0] ^ s.EKi.u[0];
		to64[1] = from64[1] ^ s.EKi.u[1];
		to64 += 2;
		from64 += 2;
	}

	to += (len - trailer);
	from += (len - trailer);

	if (trailer) {
		aes_v8_encrypt(s.aes_counter, s.EKi.c, aes_key);
		AES_INC_COUNTER(s.aes_counter);
		for (i = 0; i < trailer; i++)
			to[i] = from[i] ^ s.EKi.c[i];
	}

out:
	explicit_bzero(&s, sizeof(s));
	return (error);
}