/* crypto/bn/bn_gf2m.c */

/* ====================================================================

 * Copyright 2002 Sun Microsystems, Inc. ALL RIGHTS RESERVED.

 *

 * The Elliptic Curve Public-Key Crypto Library (ECC Code) included

 * herein is developed by SUN MICROSYSTEMS, INC., and is contributed

 * to the OpenSSL project.

 *

 * The ECC Code is licensed pursuant to the OpenSSL open source

 * license provided below.

 *

 * In addition, Sun covenants to all licensees who provide a reciprocal

 * covenant with respect to their own patents if any, not to sue under

 * current and future patent claims necessarily infringed by the making,

 * using, practicing, selling, offering for sale and/or otherwise

 * disposing of the ECC Code as delivered hereunder (or portions thereof),

 * provided that such covenant shall not apply:

 *  1) for code that a licensee deletes from the ECC Code;

 *  2) separates from the ECC Code; or

 *  3) for infringements caused by:

 *       i) the modification of the ECC Code or

 *      ii) the combination of the ECC Code with other software or

 *          devices where such combination causes the infringement.

 *

 * The software is originally written by Sheueling Chang Shantz and

 * Douglas Stebila of Sun Microsystems Laboratories.

 *

 */

/* NOTE: This file is licensed pursuant to the OpenSSL license below

 * and may be modified; but after modifications, the above covenant

 * may no longer apply!  In such cases, the corresponding paragraph

 * ["In addition, Sun covenants ... causes the infringement."] and

 * this note can be edited out; but please keep the Sun copyright

 * notice and attribution. */

/* ====================================================================

 * Copyright (c) 1998-2002 The OpenSSL Project.  All rights reserved.

 *

 * 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.

 *

 * 3. All advertising materials mentioning features or use of this

 *    software must display the following acknowledgment:

 *    "This product includes software developed by the OpenSSL Project

 *    for use in the OpenSSL Toolkit. (http://www.openssl.org/)"

 *

 * 4. The names "OpenSSL Toolkit" and "OpenSSL Project" must not be used to

 *    endorse or promote products derived from this software without

 *    prior written permission. For written permission, please contact

 *    openssl-core@openssl.org.

 *

 * 5. Products derived from this software may not be called "OpenSSL"

 *    nor may "OpenSSL" appear in their names without prior written

 *    permission of the OpenSSL Project.

 *

 * 6. Redistributions of any form whatsoever must retain the following

 *    acknowledgment:

 *    "This product includes software developed by the OpenSSL Project

 *    for use in the OpenSSL Toolkit (http://www.openssl.org/)"

 *

 * THIS SOFTWARE IS PROVIDED BY THE OpenSSL PROJECT ``AS IS'' AND ANY

 * EXPRESSED 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 OpenSSL PROJECT OR

 * ITS 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 product includes cryptographic software written by Eric Young

 * (eay@cryptsoft.com).  This product includes software written by Tim

 * Hudson (tjh@cryptsoft.com).

 *

 */

#include <assert.h>

#include <limits.h>

#include <stdio.h>

#include "cryptlib.h"

#include "bn_lcl.h"

/* Maximum number of iterations before BN_GF2m_mod_solve_quad_arr should fail. */

#define MAX_ITERATIONS 50

static const BN_ULONG SQR_tb[16] =

  {     0,     1,     4,     5,    16,    17,    20,    21,

       64,    65,    68,    69,    80,    81,    84,    85 };

/* Platform-specific macros to accelerate squaring. */

#if defined(SIXTY_FOUR_BIT) || defined(SIXTY_FOUR_BIT_LONG)

#define SQR1(w) \

    SQR_tb[(w) >> 60 & 0xF] << 56 | SQR_tb[(w) >> 56 & 0xF] << 48 | \

    SQR_tb[(w) >> 52 & 0xF] << 40 | SQR_tb[(w) >> 48 & 0xF] << 32 | \

    SQR_tb[(w) >> 44 & 0xF] << 24 | SQR_tb[(w) >> 40 & 0xF] << 16 | \

    SQR_tb[(w) >> 36 & 0xF] <<  8 | SQR_tb[(w) >> 32 & 0xF]

#define SQR0(w) \

    SQR_tb[(w) >> 28 & 0xF] << 56 | SQR_tb[(w) >> 24 & 0xF] << 48 | \

    SQR_tb[(w) >> 20 & 0xF] << 40 | SQR_tb[(w) >> 16 & 0xF] << 32 | \

    SQR_tb[(w) >> 12 & 0xF] << 24 | SQR_tb[(w) >>  8 & 0xF] << 16 | \

    SQR_tb[(w) >>  4 & 0xF] <<  8 | SQR_tb[(w)       & 0xF]

#endif

#ifdef THIRTY_TWO_BIT

#define SQR1(w) \

    SQR_tb[(w) >> 28 & 0xF] << 24 | SQR_tb[(w) >> 24 & 0xF] << 16 | \

    SQR_tb[(w) >> 20 & 0xF] <<  8 | SQR_tb[(w) >> 16 & 0xF]

#define SQR0(w) \

    SQR_tb[(w) >> 12 & 0xF] << 24 | SQR_tb[(w) >>  8 & 0xF] << 16 | \

    SQR_tb[(w) >>  4 & 0xF] <<  8 | SQR_tb[(w)       & 0xF]

#endif

#ifdef SIXTEEN_BIT

#define SQR1(w) \

    SQR_tb[(w) >> 12 & 0xF] <<  8 | SQR_tb[(w) >>  8 & 0xF]

#define SQR0(w) \

    SQR_tb[(w) >>  4 & 0xF] <<  8 | SQR_tb[(w)       & 0xF]

#endif

#ifdef EIGHT_BIT

#define SQR1(w) \

    SQR_tb[(w) >>  4 & 0xF]

#define SQR0(w) \

    SQR_tb[(w)       & 15]

#endif

/* Product of two polynomials a, b each with degree < BN_BITS2 - 1,

 * result is a polynomial r with degree < 2 * BN_BITS - 1

 * The caller MUST ensure that the variables have the right amount

 * of space allocated.

 */

#ifdef EIGHT_BIT

static void bn_GF2m_mul_1x1(BN_ULONG *r1, BN_ULONG *r0, const BN_ULONG a, const BN_ULONG b)

	{

	register BN_ULONG h, l, s;

	BN_ULONG tab[4], top1b = a >> 7;

	register BN_ULONG a1, a2;

	a1 = a & (0x7F); a2 = a1 << 1;

	tab[0] = 0; tab[1] = a1; tab[2] = a2; tab[3] = a1^a2;

	s = tab[b      & 0x3]; l  = s;

	s = tab[b >> 2 & 0x3]; l ^= s << 2; h  = s >> 6;

	s = tab[b >> 4 & 0x3]; l ^= s << 4; h ^= s >> 4;

	s = tab[b >> 6      ]; l ^= s << 6; h ^= s >> 2;

	/* compensate for the top bit of a */

	if (top1b & 01) { l ^= b << 7; h ^= b >> 1; } 

	*r1 = h; *r0 = l;

	} 

#endif

#ifdef SIXTEEN_BIT

static void bn_GF2m_mul_1x1(BN_ULONG *r1, BN_ULONG *r0, const BN_ULONG a, const BN_ULONG b)

	{

	register BN_ULONG h, l, s;

	BN_ULONG tab[4], top1b = a >> 15; 

	register BN_ULONG a1, a2;

	a1 = a & (0x7FFF); a2 = a1 << 1;

	tab[0] = 0; tab[1] = a1; tab[2] = a2; tab[3] = a1^a2;

	s = tab[b      & 0x3]; l  = s;

	s = tab[b >> 2 & 0x3]; l ^= s <<  2; h  = s >> 14;

	s = tab[b >> 4 & 0x3]; l ^= s <<  4; h ^= s >> 12;

	s = tab[b >> 6 & 0x3]; l ^= s <<  6; h ^= s >> 10;

	s = tab[b >> 8 & 0x3]; l ^= s <<  8; h ^= s >>  8;

	s = tab[b >>10 & 0x3]; l ^= s << 10; h ^= s >>  6;

	s = tab[b >>12 & 0x3]; l ^= s << 12; h ^= s >>  4;

	s = tab[b >>14      ]; l ^= s << 14; h ^= s >>  2;

	/* compensate for the top bit of a */

	if (top1b & 01) { l ^= b << 15; h ^= b >> 1; } 

	*r1 = h; *r0 = l;

	} 

#endif

#ifdef THIRTY_TWO_BIT

static void bn_GF2m_mul_1x1(BN_ULONG *r1, BN_ULONG *r0, const BN_ULONG a, const BN_ULONG b)

	{

	register BN_ULONG h, l, s;

	BN_ULONG tab[8], top2b = a >> 30; 

	register BN_ULONG a1, a2, a4;

	a1 = a & (0x3FFFFFFF); a2 = a1 << 1; a4 = a2 << 1;

	tab[0] =  0; tab[1] = a1;    tab[2] = a2;    tab[3] = a1^a2;

	tab[4] = a4; tab[5] = a1^a4; tab[6] = a2^a4; tab[7] = a1^a2^a4;

	s = tab[b       & 0x7]; l  = s;

	s = tab[b >>  3 & 0x7]; l ^= s <<  3; h  = s >> 29;

	s = tab[b >>  6 & 0x7]; l ^= s <<  6; h ^= s >> 26;

	s = tab[b >>  9 & 0x7]; l ^= s <<  9; h ^= s >> 23;

	s = tab[b >> 12 & 0x7]; l ^= s << 12; h ^= s >> 20;

	s = tab[b >> 15 & 0x7]; l ^= s << 15; h ^= s >> 17;

	s = tab[b >> 18 & 0x7]; l ^= s << 18; h ^= s >> 14;

	s = tab[b >> 21 & 0x7]; l ^= s << 21; h ^= s >> 11;

	s = tab[b >> 24 & 0x7]; l ^= s << 24; h ^= s >>  8;

	s = tab[b >> 27 & 0x7]; l ^= s << 27; h ^= s >>  5;

	s = tab[b >> 30      ]; l ^= s << 30; h ^= s >>  2;

	/* compensate for the top two bits of a */

	if (top2b & 01) { l ^= b << 30; h ^= b >> 2; } 

	if (top2b & 02) { l ^= b << 31; h ^= b >> 1; } 

	*r1 = h; *r0 = l;

	} 

#endif

#if defined(SIXTY_FOUR_BIT) || defined(SIXTY_FOUR_BIT_LONG)

static void bn_GF2m_mul_1x1(BN_ULONG *r1, BN_ULONG *r0, const BN_ULONG a, const BN_ULONG b)

	{

	register BN_ULONG h, l, s;

	BN_ULONG tab[16], top3b = a >> 61;

	register BN_ULONG a1, a2, a4, a8;

	a1 = a & (0x1FFFFFFFFFFFFFFFULL); a2 = a1 << 1; a4 = a2 << 1; a8 = a4 << 1;

	tab[ 0] = 0;     tab[ 1] = a1;       tab[ 2] = a2;       tab[ 3] = a1^a2;

	tab[ 4] = a4;    tab[ 5] = a1^a4;    tab[ 6] = a2^a4;    tab[ 7] = a1^a2^a4;

	tab[ 8] = a8;    tab[ 9] = a1^a8;    tab[10] = a2^a8;    tab[11] = a1^a2^a8;

	tab[12] = a4^a8; tab[13] = a1^a4^a8; tab[14] = a2^a4^a8; tab[15] = a1^a2^a4^a8;

	s = tab[b       & 0xF]; l  = s;

	s = tab[b >>  4 & 0xF]; l ^= s <<  4; h  = s >> 60;

	s = tab[b >>  8 & 0xF]; l ^= s <<  8; h ^= s >> 56;

	s = tab[b >> 12 & 0xF]; l ^= s << 12; h ^= s >> 52;

	s = tab[b >> 16 & 0xF]; l ^= s << 16; h ^= s >> 48;

	s = tab[b >> 20 & 0xF]; l ^= s << 20; h ^= s >> 44;

	s = tab[b >> 24 & 0xF]; l ^= s << 24; h ^= s >> 40;

	s = tab[b >> 28 & 0xF]; l ^= s << 28; h ^= s >> 36;

	s = tab[b >> 32 & 0xF]; l ^= s << 32; h ^= s >> 32;

	s = tab[b >> 36 & 0xF]; l ^= s << 36; h ^= s >> 28;

	s = tab[b >> 40 & 0xF]; l ^= s << 40; h ^= s >> 24;

	s = tab[b >> 44 & 0xF]; l ^= s << 44; h ^= s >> 20;

	s = tab[b >> 48 & 0xF]; l ^= s << 48; h ^= s >> 16;

	s = tab[b >> 52 & 0xF]; l ^= s << 52; h ^= s >> 12;

	s = tab[b >> 56 & 0xF]; l ^= s << 56; h ^= s >>  8;

	s = tab[b >> 60      ]; l ^= s << 60; h ^= s >>  4;

	/* compensate for the top three bits of a */

	if (top3b & 01) { l ^= b << 61; h ^= b >> 3; } 

	if (top3b & 02) { l ^= b << 62; h ^= b >> 2; } 

	if (top3b & 04) { l ^= b << 63; h ^= b >> 1; } 

	*r1 = h; *r0 = l;

	} 

#endif

/* Product of two polynomials a, b each with degree < 2 * BN_BITS2 - 1,

 * result is a polynomial r with degree < 4 * BN_BITS2 - 1

 * The caller MUST ensure that the variables have the right amount

 * of space allocated.

 */

static void bn_GF2m_mul_2x2(BN_ULONG *r, const BN_ULONG a1, const BN_ULONG a0, const BN_ULONG b1, const BN_ULONG b0)

	{

	BN_ULONG m1, m0;

	/* r[3] = h1, r[2] = h0; r[1] = l1; r[0] = l0 */

	bn_GF2m_mul_1x1(r+3, r+2, a1, b1);

	bn_GF2m_mul_1x1(r+1, r, a0, b0);

	bn_GF2m_mul_1x1(&m1, &m0, a0 ^ a1, b0 ^ b1);

	/* Correction on m1 ^= l1 ^ h1; m0 ^= l0 ^ h0; */

	r[2] ^= m1 ^ r[1] ^ r[3];  /* h0 ^= m1 ^ l1 ^ h1; */

	r[1] = r[3] ^ r[2] ^ r[0] ^ m1 ^ m0;  /* l1 ^= l0 ^ h0 ^ m0; */

	}

/* Add polynomials a and b and store result in r; r could be a or b, a and b 

 * could be equal; r is the bitwise XOR of a and b.

 */

EXPORT_C int	BN_GF2m_add(BIGNUM *r, const BIGNUM *a, const BIGNUM *b)

	{

	int i;

	const BIGNUM *at, *bt;

	bn_check_top(a);

	bn_check_top(b);

	if (a->top < b->top) { at = b; bt = a; }

	else { at = a; bt = b; }

	bn_wexpand(r, at->top);

	for (i = 0; i < bt->top; i++)

		{

		r->d[i] = at->d[i] ^ bt->d[i];

		}

	for (; i < at->top; i++)

		{

		r->d[i] = at->d[i];

		}

	r->top = at->top;

	bn_correct_top(r);

	return 1;

	}

/* Some functions allow for representation of the irreducible polynomials

 * as an int[], say p.  The irreducible f(t) is then of the form:

 *     t^p[0] + t^p[1] + ... + t^p[k]

 * where m = p[0] > p[1] > ... > p[k] = 0.

 */

/* Performs modular reduction of a and store result in r.  r could be a. */

EXPORT_C int BN_GF2m_mod_arr(BIGNUM *r, const BIGNUM *a, const unsigned int p[])

	{

	int j, k;

	int n, dN, d0, d1;

	BN_ULONG zz, *z;

	bn_check_top(a);

	if (!p[0])

		{

		/* reduction mod 1 => return 0 */

		BN_zero(r);

		return 1;

		}

	/* Since the algorithm does reduction in the r value, if a != r, copy

	 * the contents of a into r so we can do reduction in r. 

	 */

	if (a != r)

		{

		if (!bn_wexpand(r, a->top)) return 0;

		for (j = 0; j < a->top; j++)

			{

			r->d[j] = a->d[j];

			}

		r->top = a->top;

		}

	z = r->d;

	/* start reduction */

	dN = p[0] / BN_BITS2;  

	for (j = r->top - 1; j > dN;)

		{

		zz = z[j];

		if (z[j] == 0) { j--; continue; }

		z[j] = 0;

		for (k = 1; p[k] != 0; k++)

			{

			/* reducing component t^p[k] */

			n = p[0] - p[k];

			d0 = n % BN_BITS2;  d1 = BN_BITS2 - d0;

			n /= BN_BITS2; 

			z[j-n] ^= (zz>>d0);

			if (d0) z[j-n-1] ^= (zz<<d1);

			}

		/* reducing component t^0 */

		n = dN;  

		d0 = p[0] % BN_BITS2;

		d1 = BN_BITS2 - d0;

		z[j-n] ^= (zz >> d0);

		if (d0) z[j-n-1] ^= (zz << d1);

		}

	/* final round of reduction */

	while (j == dN)

		{

		d0 = p[0] % BN_BITS2;

		zz = z[dN] >> d0;

		if (zz == 0) break;

		d1 = BN_BITS2 - d0;

		if (d0) z[dN] = (z[dN] << d1) >> d1; /* clear up the top d1 bits */

		z[0] ^= zz; /* reduction t^0 component */

		for (k = 1; p[k] != 0; k++)

			{

			BN_ULONG tmp_ulong;

			/* reducing component t^p[k]*/

			n = p[k] / BN_BITS2;   

			d0 = p[k] % BN_BITS2;

			d1 = BN_BITS2 - d0;

			z[n] ^= (zz << d0);

			tmp_ulong = zz >> d1;

                        if (d0 && tmp_ulong)

                                z[n+1] ^= tmp_ulong;

			}

		}

	bn_correct_top(r);

	return 1;

	}

/* Performs modular reduction of a by p and store result in r.  r could be a.

 *

 * This function calls down to the BN_GF2m_mod_arr implementation; this wrapper

 * function is only provided for convenience; for best performance, use the 

 * BN_GF2m_mod_arr function.

 */

EXPORT_C int	BN_GF2m_mod(BIGNUM *r, const BIGNUM *a, const BIGNUM *p)

	{

	int ret = 0;

	const int max = BN_num_bits(p);

	unsigned int *arr=NULL;

	bn_check_top(a);

	bn_check_top(p);

	if ((arr = (unsigned int *)OPENSSL_malloc(sizeof(unsigned int) * max)) == NULL) goto err;

	ret = BN_GF2m_poly2arr(p, arr, max);

	if (!ret || ret > max)

		{

		BNerr(BN_F_BN_GF2M_MOD,BN_R_INVALID_LENGTH);

		goto err;

		}

	ret = BN_GF2m_mod_arr(r, a, arr);

	bn_check_top(r);

err:

	if (arr) OPENSSL_free(arr);

	return ret;

	}

/* Compute the product of two polynomials a and b, reduce modulo p, and store

 * the result in r.  r could be a or b; a could be b.

 */

EXPORT_C int	BN_GF2m_mod_mul_arr(BIGNUM *r, const BIGNUM *a, const BIGNUM *b, const unsigned int p[], BN_CTX *ctx)

	{

	int zlen, i, j, k, ret = 0;

	BIGNUM *s;

	BN_ULONG x1, x0, y1, y0, zz[4];

	bn_check_top(a);

	bn_check_top(b);

	if (a == b)

		{

		return BN_GF2m_mod_sqr_arr(r, a, p, ctx);

		}

	BN_CTX_start(ctx);

	if ((s = BN_CTX_get(ctx)) == NULL) goto err;

	zlen = a->top + b->top + 4;

	if (!bn_wexpand(s, zlen)) goto err;

	s->top = zlen;

	for (i = 0; i < zlen; i++) s->d[i] = 0;

	for (j = 0; j < b->top; j += 2)

		{

		y0 = b->d[j];

		y1 = ((j+1) == b->top) ? 0 : b->d[j+1];

		for (i = 0; i < a->top; i += 2)

			{

			x0 = a->d[i];

			x1 = ((i+1) == a->top) ? 0 : a->d[i+1];

			bn_GF2m_mul_2x2(zz, x1, x0, y1, y0);

			for (k = 0; k < 4; k++) s->d[i+j+k] ^= zz[k];

			}

		}

	bn_correct_top(s);

	if (BN_GF2m_mod_arr(r, s, p))

		ret = 1;

	bn_check_top(r);

err:

	BN_CTX_end(ctx);

	return ret;

	}

/* Compute the product of two polynomials a and b, reduce modulo p, and store

 * the result in r.  r could be a or b; a could equal b.

 *

 * This function calls down to the BN_GF2m_mod_mul_arr implementation; this wrapper

 * function is only provided for convenience; for best performance, use the 

 * BN_GF2m_mod_mul_arr function.

 */

EXPORT_C int	BN_GF2m_mod_mul(BIGNUM *r, const BIGNUM *a, const BIGNUM *b, const BIGNUM *p, BN_CTX *ctx)

	{

	int ret = 0;

	const int max = BN_num_bits(p);

	unsigned int *arr=NULL;

	bn_check_top(a);

	bn_check_top(b);

	bn_check_top(p);

	if ((arr = (unsigned int *)OPENSSL_malloc(sizeof(unsigned int) * max)) == NULL) goto err;

	ret = BN_GF2m_poly2arr(p, arr, max);

	if (!ret || ret > max)

		{

		BNerr(BN_F_BN_GF2M_MOD_MUL,BN_R_INVALID_LENGTH);

		goto err;

		}

	ret = BN_GF2m_mod_mul_arr(r, a, b, arr, ctx);

	bn_check_top(r);

err:

	if (arr) OPENSSL_free(arr);

	return ret;

	}

/* Square a, reduce the result mod p, and store it in a.  r could be a. */

EXPORT_C int	BN_GF2m_mod_sqr_arr(BIGNUM *r, const BIGNUM *a, const unsigned int p[], BN_CTX *ctx)

	{

	int i, ret = 0;

	BIGNUM *s;

	bn_check_top(a);

	BN_CTX_start(ctx);

	if ((s = BN_CTX_get(ctx)) == NULL) return 0;

	if (!bn_wexpand(s, 2 * a->top)) goto err;