vmac.c 18.9 KB
Newer Older
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
/*
 * Modified to interface to the Linux kernel
 * Copyright (c) 2009, Intel Corporation.
 *
 * This program is free software; you can redistribute it and/or modify it
 * under the terms and conditions of the GNU General Public License,
 * version 2, as published by the Free Software Foundation.
 *
 * This program is distributed in the hope it will be useful, but WITHOUT
 * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
 * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
 * more details.
 *
 * You should have received a copy of the GNU General Public License along with
 * this program; if not, write to the Free Software Foundation, Inc., 59 Temple
 * Place - Suite 330, Boston, MA 02111-1307 USA.
 */

/* --------------------------------------------------------------------------
 * VMAC and VHASH Implementation by Ted Krovetz (tdk@acm.org) and Wei Dai.
 * This implementation is herby placed in the public domain.
 * The authors offers no warranty. Use at your own risk.
 * Please send bug reports to the authors.
 * Last modified: 17 APR 08, 1700 PDT
 * ----------------------------------------------------------------------- */

#include <linux/init.h>
#include <linux/types.h>
#include <linux/crypto.h>
30
#include <linux/module.h>
31 32 33 34 35 36 37 38 39 40
#include <linux/scatterlist.h>
#include <asm/byteorder.h>
#include <crypto/scatterwalk.h>
#include <crypto/vmac.h>
#include <crypto/internal/hash.h>

/*
 * Constants and masks
 */
#define UINT64_C(x) x##ULL
41 42 43 44 45
static const u64 p64   = UINT64_C(0xfffffffffffffeff);	/* 2^64 - 257 prime  */
static const u64 m62   = UINT64_C(0x3fffffffffffffff);	/* 62-bit mask       */
static const u64 m63   = UINT64_C(0x7fffffffffffffff);	/* 63-bit mask       */
static const u64 m64   = UINT64_C(0xffffffffffffffff);	/* 64-bit mask       */
static const u64 mpoly = UINT64_C(0x1fffffff1fffffff);	/* Poly key mask     */
46

47 48
#define pe64_to_cpup le64_to_cpup		/* Prefer little endian */

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
#ifdef __LITTLE_ENDIAN
#define INDEX_HIGH 1
#define INDEX_LOW 0
#else
#define INDEX_HIGH 0
#define INDEX_LOW 1
#endif

/*
 * The following routines are used in this implementation. They are
 * written via macros to simulate zero-overhead call-by-reference.
 *
 * MUL64: 64x64->128-bit multiplication
 * PMUL64: assumes top bits cleared on inputs
 * ADD128: 128x128->128-bit addition
 */

#define ADD128(rh, rl, ih, il)						\
	do {								\
		u64 _il = (il);						\
		(rl) += (_il);						\
		if ((rl) < (_il))					\
			(rh)++;						\
		(rh) += (ih);						\
	} while (0)

#define MUL32(i1, i2)	((u64)(u32)(i1)*(u32)(i2))

#define PMUL64(rh, rl, i1, i2)	/* Assumes m doesn't overflow */	\
	do {								\
		u64 _i1 = (i1), _i2 = (i2);				\
		u64 m = MUL32(_i1, _i2>>32) + MUL32(_i1>>32, _i2);	\
		rh = MUL32(_i1>>32, _i2>>32);				\
		rl = MUL32(_i1, _i2);					\
		ADD128(rh, rl, (m >> 32), (m << 32));			\
	} while (0)

#define MUL64(rh, rl, i1, i2)						\
	do {								\
		u64 _i1 = (i1), _i2 = (i2);				\
		u64 m1 = MUL32(_i1, _i2>>32);				\
		u64 m2 = MUL32(_i1>>32, _i2);				\
		rh = MUL32(_i1>>32, _i2>>32);				\
		rl = MUL32(_i1, _i2);					\
		ADD128(rh, rl, (m1 >> 32), (m1 << 32));			\
		ADD128(rh, rl, (m2 >> 32), (m2 << 32));			\
	} while (0)

/*
 * For highest performance the L1 NH and L2 polynomial hashes should be
Lucas De Marchi's avatar
Lucas De Marchi committed
99
 * carefully implemented to take advantage of one's target architecture.
100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115
 * Here these two hash functions are defined multiple time; once for
 * 64-bit architectures, once for 32-bit SSE2 architectures, and once
 * for the rest (32-bit) architectures.
 * For each, nh_16 *must* be defined (works on multiples of 16 bytes).
 * Optionally, nh_vmac_nhbytes can be defined (for multiples of
 * VMAC_NHBYTES), and nh_16_2 and nh_vmac_nhbytes_2 (versions that do two
 * NH computations at once).
 */

#ifdef CONFIG_64BIT

#define nh_16(mp, kp, nw, rh, rl)					\
	do {								\
		int i; u64 th, tl;					\
		rh = rl = 0;						\
		for (i = 0; i < nw; i += 2) {				\
116 117
			MUL64(th, tl, pe64_to_cpup((mp)+i)+(kp)[i],	\
				pe64_to_cpup((mp)+i+1)+(kp)[i+1]);	\
118 119 120 121 122 123 124 125 126
			ADD128(rh, rl, th, tl);				\
		}							\
	} while (0)

#define nh_16_2(mp, kp, nw, rh, rl, rh1, rl1)				\
	do {								\
		int i; u64 th, tl;					\
		rh1 = rl1 = rh = rl = 0;				\
		for (i = 0; i < nw; i += 2) {				\
127 128
			MUL64(th, tl, pe64_to_cpup((mp)+i)+(kp)[i],	\
				pe64_to_cpup((mp)+i+1)+(kp)[i+1]);	\
129
			ADD128(rh, rl, th, tl);				\
130 131
			MUL64(th, tl, pe64_to_cpup((mp)+i)+(kp)[i+2],	\
				pe64_to_cpup((mp)+i+1)+(kp)[i+3]);	\
132 133 134 135 136 137 138 139 140 141
			ADD128(rh1, rl1, th, tl);			\
		}							\
	} while (0)

#if (VMAC_NHBYTES >= 64) /* These versions do 64-bytes of message at a time */
#define nh_vmac_nhbytes(mp, kp, nw, rh, rl)				\
	do {								\
		int i; u64 th, tl;					\
		rh = rl = 0;						\
		for (i = 0; i < nw; i += 8) {				\
142 143
			MUL64(th, tl, pe64_to_cpup((mp)+i)+(kp)[i],	\
				pe64_to_cpup((mp)+i+1)+(kp)[i+1]);	\
144
			ADD128(rh, rl, th, tl);				\
145 146
			MUL64(th, tl, pe64_to_cpup((mp)+i+2)+(kp)[i+2],	\
				pe64_to_cpup((mp)+i+3)+(kp)[i+3]);	\
147
			ADD128(rh, rl, th, tl);				\
148 149
			MUL64(th, tl, pe64_to_cpup((mp)+i+4)+(kp)[i+4],	\
				pe64_to_cpup((mp)+i+5)+(kp)[i+5]);	\
150
			ADD128(rh, rl, th, tl);				\
151 152
			MUL64(th, tl, pe64_to_cpup((mp)+i+6)+(kp)[i+6],	\
				pe64_to_cpup((mp)+i+7)+(kp)[i+7]);	\
153 154 155 156 157 158 159 160 161
			ADD128(rh, rl, th, tl);				\
		}							\
	} while (0)

#define nh_vmac_nhbytes_2(mp, kp, nw, rh, rl, rh1, rl1)			\
	do {								\
		int i; u64 th, tl;					\
		rh1 = rl1 = rh = rl = 0;				\
		for (i = 0; i < nw; i += 8) {				\
162 163
			MUL64(th, tl, pe64_to_cpup((mp)+i)+(kp)[i],	\
				pe64_to_cpup((mp)+i+1)+(kp)[i+1]);	\
164
			ADD128(rh, rl, th, tl);				\
165 166
			MUL64(th, tl, pe64_to_cpup((mp)+i)+(kp)[i+2],	\
				pe64_to_cpup((mp)+i+1)+(kp)[i+3]);	\
167
			ADD128(rh1, rl1, th, tl);			\
168 169
			MUL64(th, tl, pe64_to_cpup((mp)+i+2)+(kp)[i+2],	\
				pe64_to_cpup((mp)+i+3)+(kp)[i+3]);	\
170
			ADD128(rh, rl, th, tl);				\
171 172
			MUL64(th, tl, pe64_to_cpup((mp)+i+2)+(kp)[i+4],	\
				pe64_to_cpup((mp)+i+3)+(kp)[i+5]);	\
173
			ADD128(rh1, rl1, th, tl);			\
174 175
			MUL64(th, tl, pe64_to_cpup((mp)+i+4)+(kp)[i+4],	\
				pe64_to_cpup((mp)+i+5)+(kp)[i+5]);	\
176
			ADD128(rh, rl, th, tl);				\
177 178
			MUL64(th, tl, pe64_to_cpup((mp)+i+4)+(kp)[i+6],	\
				pe64_to_cpup((mp)+i+5)+(kp)[i+7]);	\
179
			ADD128(rh1, rl1, th, tl);			\
180 181
			MUL64(th, tl, pe64_to_cpup((mp)+i+6)+(kp)[i+6],	\
				pe64_to_cpup((mp)+i+7)+(kp)[i+7]);	\
182
			ADD128(rh, rl, th, tl);				\
183 184
			MUL64(th, tl, pe64_to_cpup((mp)+i+6)+(kp)[i+8],	\
				pe64_to_cpup((mp)+i+7)+(kp)[i+9]);	\
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
			ADD128(rh1, rl1, th, tl);			\
		}							\
	} while (0)
#endif

#define poly_step(ah, al, kh, kl, mh, ml)				\
	do {								\
		u64 t1h, t1l, t2h, t2l, t3h, t3l, z = 0;		\
		/* compute ab*cd, put bd into result registers */	\
		PMUL64(t3h, t3l, al, kh);				\
		PMUL64(t2h, t2l, ah, kl);				\
		PMUL64(t1h, t1l, ah, 2*kh);				\
		PMUL64(ah, al, al, kl);					\
		/* add 2 * ac to result */				\
		ADD128(ah, al, t1h, t1l);				\
		/* add together ad + bc */				\
		ADD128(t2h, t2l, t3h, t3l);				\
		/* now (ah,al), (t2l,2*t2h) need summing */		\
		/* first add the high registers, carrying into t2h */	\
		ADD128(t2h, ah, z, t2l);				\
		/* double t2h and add top bit of ah */			\
		t2h = 2 * t2h + (ah >> 63);				\
		ah &= m63;						\
		/* now add the low registers */				\
		ADD128(ah, al, mh, ml);					\
		ADD128(ah, al, z, t2h);					\
	} while (0)

#else /* ! CONFIG_64BIT */

#ifndef nh_16
#define nh_16(mp, kp, nw, rh, rl)					\
	do {								\
		u64 t1, t2, m1, m2, t;					\
		int i;							\
		rh = rl = t = 0;					\
		for (i = 0; i < nw; i += 2)  {				\
222 223
			t1 = pe64_to_cpup(mp+i) + kp[i];		\
			t2 = pe64_to_cpup(mp+i+1) + kp[i+1];		\
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
			m2 = MUL32(t1 >> 32, t2);			\
			m1 = MUL32(t1, t2 >> 32);			\
			ADD128(rh, rl, MUL32(t1 >> 32, t2 >> 32),	\
				MUL32(t1, t2));				\
			rh += (u64)(u32)(m1 >> 32)			\
				+ (u32)(m2 >> 32);			\
			t += (u64)(u32)m1 + (u32)m2;			\
		}							\
		ADD128(rh, rl, (t >> 32), (t << 32));			\
	} while (0)
#endif

static void poly_step_func(u64 *ahi, u64 *alo,
			const u64 *kh, const u64 *kl,
			const u64 *mh, const u64 *ml)
{
#define a0 (*(((u32 *)alo)+INDEX_LOW))
#define a1 (*(((u32 *)alo)+INDEX_HIGH))
#define a2 (*(((u32 *)ahi)+INDEX_LOW))
#define a3 (*(((u32 *)ahi)+INDEX_HIGH))
#define k0 (*(((u32 *)kl)+INDEX_LOW))
#define k1 (*(((u32 *)kl)+INDEX_HIGH))
#define k2 (*(((u32 *)kh)+INDEX_LOW))
#define k3 (*(((u32 *)kh)+INDEX_HIGH))

	u64 p, q, t;
	u32 t2;

	p = MUL32(a3, k3);
	p += p;
	p += *(u64 *)mh;
	p += MUL32(a0, k2);
	p += MUL32(a1, k1);
	p += MUL32(a2, k0);
	t = (u32)(p);
	p >>= 32;
	p += MUL32(a0, k3);
	p += MUL32(a1, k2);
	p += MUL32(a2, k1);
	p += MUL32(a3, k0);
	t |= ((u64)((u32)p & 0x7fffffff)) << 32;
	p >>= 31;
	p += (u64)(((u32 *)ml)[INDEX_LOW]);
	p += MUL32(a0, k0);
	q =  MUL32(a1, k3);
	q += MUL32(a2, k2);
	q += MUL32(a3, k1);
	q += q;
	p += q;
	t2 = (u32)(p);
	p >>= 32;
	p += (u64)(((u32 *)ml)[INDEX_HIGH]);
	p += MUL32(a0, k1);
	p += MUL32(a1, k0);
	q =  MUL32(a2, k3);
	q += MUL32(a3, k2);
	q += q;
	p += q;
	*(u64 *)(alo) = (p << 32) | t2;
	p >>= 32;
	*(u64 *)(ahi) = p + t;

#undef a0
#undef a1
#undef a2
#undef a3
#undef k0
#undef k1
#undef k2
#undef k3
}

#define poly_step(ah, al, kh, kl, mh, ml)				\
	poly_step_func(&(ah), &(al), &(kh), &(kl), &(mh), &(ml))

#endif  /* end of specialized NH and poly definitions */

/* At least nh_16 is defined. Defined others as needed here */
#ifndef nh_16_2
#define nh_16_2(mp, kp, nw, rh, rl, rh2, rl2)				\
	do { 								\
		nh_16(mp, kp, nw, rh, rl);				\
		nh_16(mp, ((kp)+2), nw, rh2, rl2);			\
	} while (0)
#endif
#ifndef nh_vmac_nhbytes
#define nh_vmac_nhbytes(mp, kp, nw, rh, rl)				\
	nh_16(mp, kp, nw, rh, rl)
#endif
#ifndef nh_vmac_nhbytes_2
#define nh_vmac_nhbytes_2(mp, kp, nw, rh, rl, rh2, rl2)			\
	do {								\
		nh_vmac_nhbytes(mp, kp, nw, rh, rl);			\
		nh_vmac_nhbytes(mp, ((kp)+2), nw, rh2, rl2);		\
	} while (0)
#endif

static void vhash_abort(struct vmac_ctx *ctx)
{
	ctx->polytmp[0] = ctx->polykey[0] ;
	ctx->polytmp[1] = ctx->polykey[1] ;
	ctx->first_block_processed = 0;
}

328
static u64 l3hash(u64 p1, u64 p2, u64 k1, u64 k2, u64 len)
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
{
	u64 rh, rl, t, z = 0;

	/* fully reduce (p1,p2)+(len,0) mod p127 */
	t = p1 >> 63;
	p1 &= m63;
	ADD128(p1, p2, len, t);
	/* At this point, (p1,p2) is at most 2^127+(len<<64) */
	t = (p1 > m63) + ((p1 == m63) && (p2 == m64));
	ADD128(p1, p2, z, t);
	p1 &= m63;

	/* compute (p1,p2)/(2^64-2^32) and (p1,p2)%(2^64-2^32) */
	t = p1 + (p2 >> 32);
	t += (t >> 32);
	t += (u32)t > 0xfffffffeu;
	p1 += (t >> 32);
	p2 += (p1 << 32);

	/* compute (p1+k1)%p64 and (p2+k2)%p64 */
	p1 += k1;
	p1 += (0 - (p1 < k1)) & 257;
	p2 += k2;
	p2 += (0 - (p2 < k2)) & 257;

	/* compute (p1+k1)*(p2+k2)%p64 */
	MUL64(rh, rl, p1, p2);
	t = rh >> 56;
	ADD128(t, rl, z, rh);
	rh <<= 8;
	ADD128(t, rl, z, rh);
	t += t << 8;
	rl += t;
	rl += (0 - (rl < t)) & 257;
	rl += (0 - (rl > p64-1)) & 257;
	return rl;
}

static void vhash_update(const unsigned char *m,
			unsigned int mbytes, /* Pos multiple of VMAC_NHBYTES */
			struct vmac_ctx *ctx)
{
	u64 rh, rl, *mptr;
	const u64 *kptr = (u64 *)ctx->nhkey;
	int i;
	u64 ch, cl;
	u64 pkh = ctx->polykey[0];
	u64 pkl = ctx->polykey[1];

378 379 380 381 382
	if (!mbytes)
		return;

	BUG_ON(mbytes % VMAC_NHBYTES);

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
	mptr = (u64 *)m;
	i = mbytes / VMAC_NHBYTES;  /* Must be non-zero */

	ch = ctx->polytmp[0];
	cl = ctx->polytmp[1];

	if (!ctx->first_block_processed) {
		ctx->first_block_processed = 1;
		nh_vmac_nhbytes(mptr, kptr, VMAC_NHBYTES/8, rh, rl);
		rh &= m62;
		ADD128(ch, cl, rh, rl);
		mptr += (VMAC_NHBYTES/sizeof(u64));
		i--;
	}

	while (i--) {
		nh_vmac_nhbytes(mptr, kptr, VMAC_NHBYTES/8, rh, rl);
		rh &= m62;
		poly_step(ch, cl, pkh, pkl, rh, rl);
		mptr += (VMAC_NHBYTES/sizeof(u64));
	}

	ctx->polytmp[0] = ch;
	ctx->polytmp[1] = cl;
}

static u64 vhash(unsigned char m[], unsigned int mbytes,
			u64 *tagl, struct vmac_ctx *ctx)
{
	u64 rh, rl, *mptr;
	const u64 *kptr = (u64 *)ctx->nhkey;
	int i, remaining;
	u64 ch, cl;
	u64 pkh = ctx->polykey[0];
	u64 pkl = ctx->polykey[1];

	mptr = (u64 *)m;
	i = mbytes / VMAC_NHBYTES;
	remaining = mbytes % VMAC_NHBYTES;

	if (ctx->first_block_processed) {
		ch = ctx->polytmp[0];
		cl = ctx->polytmp[1];
	} else if (i) {
		nh_vmac_nhbytes(mptr, kptr, VMAC_NHBYTES/8, ch, cl);
		ch &= m62;
		ADD128(ch, cl, pkh, pkl);
		mptr += (VMAC_NHBYTES/sizeof(u64));
		i--;
	} else if (remaining) {
		nh_16(mptr, kptr, 2*((remaining+15)/16), ch, cl);
		ch &= m62;
		ADD128(ch, cl, pkh, pkl);
		mptr += (VMAC_NHBYTES/sizeof(u64));
		goto do_l3;
	} else {/* Empty String */
		ch = pkh; cl = pkl;
		goto do_l3;
	}

	while (i--) {
		nh_vmac_nhbytes(mptr, kptr, VMAC_NHBYTES/8, rh, rl);
		rh &= m62;
		poly_step(ch, cl, pkh, pkl, rh, rl);
		mptr += (VMAC_NHBYTES/sizeof(u64));
	}
	if (remaining) {
		nh_16(mptr, kptr, 2*((remaining+15)/16), rh, rl);
		rh &= m62;
		poly_step(ch, cl, pkh, pkl, rh, rl);
	}

do_l3:
	vhash_abort(ctx);
	remaining *= 8;
	return l3hash(ch, cl, ctx->l3key[0], ctx->l3key[1], remaining);
}

static u64 vmac(unsigned char m[], unsigned int mbytes,
462
			const unsigned char n[16], u64 *tagl,
463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483
			struct vmac_ctx_t *ctx)
{
	u64 *in_n, *out_p;
	u64 p, h;
	int i;

	in_n = ctx->__vmac_ctx.cached_nonce;
	out_p = ctx->__vmac_ctx.cached_aes;

	i = n[15] & 1;
	if ((*(u64 *)(n+8) != in_n[1]) || (*(u64 *)(n) != in_n[0])) {
		in_n[0] = *(u64 *)(n);
		in_n[1] = *(u64 *)(n+8);
		((unsigned char *)in_n)[15] &= 0xFE;
		crypto_cipher_encrypt_one(ctx->child,
			(unsigned char *)out_p, (unsigned char *)in_n);

		((unsigned char *)in_n)[15] |= (unsigned char)(1-i);
	}
	p = be64_to_cpup(out_p + i);
	h = vhash(m, mbytes, (u64 *)0, &ctx->__vmac_ctx);
484
	return le64_to_cpu(p + h);
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
}

static int vmac_set_key(unsigned char user_key[], struct vmac_ctx_t *ctx)
{
	u64 in[2] = {0}, out[2];
	unsigned i;
	int err = 0;

	err = crypto_cipher_setkey(ctx->child, user_key, VMAC_KEY_LEN);
	if (err)
		return err;

	/* Fill nh key */
	((unsigned char *)in)[0] = 0x80;
	for (i = 0; i < sizeof(ctx->__vmac_ctx.nhkey)/8; i += 2) {
		crypto_cipher_encrypt_one(ctx->child,
			(unsigned char *)out, (unsigned char *)in);
		ctx->__vmac_ctx.nhkey[i] = be64_to_cpup(out);
		ctx->__vmac_ctx.nhkey[i+1] = be64_to_cpup(out+1);
		((unsigned char *)in)[15] += 1;
	}

	/* Fill poly key */
	((unsigned char *)in)[0] = 0xC0;
	in[1] = 0;
	for (i = 0; i < sizeof(ctx->__vmac_ctx.polykey)/8; i += 2) {
		crypto_cipher_encrypt_one(ctx->child,
			(unsigned char *)out, (unsigned char *)in);
		ctx->__vmac_ctx.polytmp[i] =
			ctx->__vmac_ctx.polykey[i] =
				be64_to_cpup(out) & mpoly;
		ctx->__vmac_ctx.polytmp[i+1] =
			ctx->__vmac_ctx.polykey[i+1] =
				be64_to_cpup(out+1) & mpoly;
		((unsigned char *)in)[15] += 1;
	}

	/* Fill ip key */
	((unsigned char *)in)[0] = 0xE0;
	in[1] = 0;
	for (i = 0; i < sizeof(ctx->__vmac_ctx.l3key)/8; i += 2) {
		do {
			crypto_cipher_encrypt_one(ctx->child,
				(unsigned char *)out, (unsigned char *)in);
			ctx->__vmac_ctx.l3key[i] = be64_to_cpup(out);
			ctx->__vmac_ctx.l3key[i+1] = be64_to_cpup(out+1);
			((unsigned char *)in)[15] += 1;
		} while (ctx->__vmac_ctx.l3key[i] >= p64
			|| ctx->__vmac_ctx.l3key[i+1] >= p64);
	}

	/* Invalidate nonce/aes cache and reset other elements */
	ctx->__vmac_ctx.cached_nonce[0] = (u64)-1; /* Ensure illegal nonce */
	ctx->__vmac_ctx.cached_nonce[1] = (u64)0;  /* Ensure illegal nonce */
	ctx->__vmac_ctx.first_block_processed = 0;

	return err;
}

static int vmac_setkey(struct crypto_shash *parent,
		const u8 *key, unsigned int keylen)
{
	struct vmac_ctx_t *ctx = crypto_shash_ctx(parent);

	if (keylen != VMAC_KEY_LEN) {
		crypto_shash_set_flags(parent, CRYPTO_TFM_RES_BAD_KEY_LEN);
		return -EINVAL;
	}

	return vmac_set_key((u8 *)key, ctx);
}

static int vmac_init(struct shash_desc *pdesc)
{
	return 0;
}

static int vmac_update(struct shash_desc *pdesc, const u8 *p,
		unsigned int len)
{
	struct crypto_shash *parent = pdesc->tfm;
	struct vmac_ctx_t *ctx = crypto_shash_ctx(parent);
567 568 569 570 571 572 573 574 575 576 577 578 579
	int expand;
	int min;

	expand = VMAC_NHBYTES - ctx->partial_size > 0 ?
			VMAC_NHBYTES - ctx->partial_size : 0;

	min = len < expand ? len : expand;

	memcpy(ctx->partial + ctx->partial_size, p, min);
	ctx->partial_size += min;

	if (len < expand)
		return 0;
580

581 582 583 584 585 586 587 588 589 590 591 592 593
	vhash_update(ctx->partial, VMAC_NHBYTES, &ctx->__vmac_ctx);
	ctx->partial_size = 0;

	len -= expand;
	p += expand;

	if (len % VMAC_NHBYTES) {
		memcpy(ctx->partial, p + len - (len % VMAC_NHBYTES),
			len % VMAC_NHBYTES);
		ctx->partial_size = len % VMAC_NHBYTES;
	}

	vhash_update(p, len - len % VMAC_NHBYTES, &ctx->__vmac_ctx);
594 595 596 597 598 599 600 601 602 603 604

	return 0;
}

static int vmac_final(struct shash_desc *pdesc, u8 *out)
{
	struct crypto_shash *parent = pdesc->tfm;
	struct vmac_ctx_t *ctx = crypto_shash_ctx(parent);
	vmac_t mac;
	u8 nonce[16] = {};

605 606 607 608 609 610 611 612 613 614
	/* vmac() ends up accessing outside the array bounds that
	 * we specify.  In appears to access up to the next 2-word
	 * boundary.  We'll just be uber cautious and zero the
	 * unwritten bytes in the buffer.
	 */
	if (ctx->partial_size) {
		memset(ctx->partial + ctx->partial_size, 0,
			VMAC_NHBYTES - ctx->partial_size);
	}
	mac = vmac(ctx->partial, ctx->partial_size, nonce, NULL, ctx);
615
	memcpy(out, &mac, sizeof(vmac_t));
616
	memzero_explicit(&mac, sizeof(vmac_t));
617
	memset(&ctx->__vmac_ctx, 0, sizeof(struct vmac_ctx));
618
	ctx->partial_size = 0;
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
	return 0;
}

static int vmac_init_tfm(struct crypto_tfm *tfm)
{
	struct crypto_cipher *cipher;
	struct crypto_instance *inst = (void *)tfm->__crt_alg;
	struct crypto_spawn *spawn = crypto_instance_ctx(inst);
	struct vmac_ctx_t *ctx = crypto_tfm_ctx(tfm);

	cipher = crypto_spawn_cipher(spawn);
	if (IS_ERR(cipher))
		return PTR_ERR(cipher);

	ctx->child = cipher;
	return 0;
}

static void vmac_exit_tfm(struct crypto_tfm *tfm)
{
	struct vmac_ctx_t *ctx = crypto_tfm_ctx(tfm);
	crypto_free_cipher(ctx->child);
}

static int vmac_create(struct crypto_template *tmpl, struct rtattr **tb)
{
	struct shash_instance *inst;
	struct crypto_alg *alg;
	int err;

	err = crypto_check_attr_type(tb, CRYPTO_ALG_TYPE_SHASH);
	if (err)
		return err;

	alg = crypto_get_attr_alg(tb, CRYPTO_ALG_TYPE_CIPHER,
			CRYPTO_ALG_TYPE_MASK);
	if (IS_ERR(alg))
		return PTR_ERR(alg);

	inst = shash_alloc_instance("vmac", alg);
	err = PTR_ERR(inst);
	if (IS_ERR(inst))
		goto out_put_alg;

	err = crypto_init_spawn(shash_instance_ctx(inst), alg,
			shash_crypto_instance(inst),
			CRYPTO_ALG_TYPE_MASK);
	if (err)
		goto out_free_inst;

	inst->alg.base.cra_priority = alg->cra_priority;
	inst->alg.base.cra_blocksize = alg->cra_blocksize;
	inst->alg.base.cra_alignmask = alg->cra_alignmask;

	inst->alg.digestsize = sizeof(vmac_t);
	inst->alg.base.cra_ctxsize = sizeof(struct vmac_ctx_t);
	inst->alg.base.cra_init = vmac_init_tfm;
	inst->alg.base.cra_exit = vmac_exit_tfm;

	inst->alg.init = vmac_init;
	inst->alg.update = vmac_update;
	inst->alg.final = vmac_final;
	inst->alg.setkey = vmac_setkey;

	err = shash_register_instance(tmpl, inst);
	if (err) {
out_free_inst:
		shash_free_instance(shash_crypto_instance(inst));
	}

out_put_alg:
	crypto_mod_put(alg);
	return err;
}

static struct crypto_template vmac_tmpl = {
	.name = "vmac",
	.create = vmac_create,
	.free = shash_free_instance,
	.module = THIS_MODULE,
};

static int __init vmac_module_init(void)
{
	return crypto_register_template(&vmac_tmpl);
}

static void __exit vmac_module_exit(void)
{
	crypto_unregister_template(&vmac_tmpl);
}

module_init(vmac_module_init);
module_exit(vmac_module_exit);

MODULE_LICENSE("GPL");
MODULE_DESCRIPTION("VMAC hash algorithm");
716
MODULE_ALIAS_CRYPTO("vmac");