aes_x86core.c

OpenSSL 0.9.8k 最新版OpenSSL

    U64(0x694b775a694b775a), U64(0x161a121c161a121c),
    U64(0x0aba93e20aba93e2), U64(0xe52aa0c0e52aa0c0),
    U64(0x43e0223c43e0223c), U64(0x1d171b121d171b12),
    U64(0x0b0d090e0b0d090e), U64(0xadc78bf2adc78bf2),
    U64(0xb9a8b62db9a8b62d), U64(0xc8a91e14c8a91e14),
    U64(0x8519f1578519f157), U64(0x4c0775af4c0775af),
    U64(0xbbdd99eebbdd99ee), U64(0xfd607fa3fd607fa3),
    U64(0x9f2601f79f2601f7), U64(0xbcf5725cbcf5725c),
    U64(0xc53b6644c53b6644), U64(0x347efb5b347efb5b),
    U64(0x7629438b7629438b), U64(0xdcc623cbdcc623cb),
    U64(0x68fcedb668fcedb6), U64(0x63f1e4b863f1e4b8),
    U64(0xcadc31d7cadc31d7), U64(0x1085634210856342),
    U64(0x4022971340229713), U64(0x2011c6842011c684),
    U64(0x7d244a857d244a85), U64(0xf83dbbd2f83dbbd2),
    U64(0x1132f9ae1132f9ae), U64(0x6da129c76da129c7),
    U64(0x4b2f9e1d4b2f9e1d), U64(0xf330b2dcf330b2dc),
    U64(0xec52860dec52860d), U64(0xd0e3c177d0e3c177),
    U64(0x6c16b32b6c16b32b), U64(0x99b970a999b970a9),
    U64(0xfa489411fa489411), U64(0x2264e9472264e947),
    U64(0xc48cfca8c48cfca8), U64(0x1a3ff0a01a3ff0a0),
    U64(0xd82c7d56d82c7d56), U64(0xef903322ef903322),
    U64(0xc74e4987c74e4987), U64(0xc1d138d9c1d138d9),
    U64(0xfea2ca8cfea2ca8c), U64(0x360bd498360bd498),
    U64(0xcf81f5a6cf81f5a6), U64(0x28de7aa528de7aa5),
    U64(0x268eb7da268eb7da), U64(0xa4bfad3fa4bfad3f),
    U64(0xe49d3a2ce49d3a2c), U64(0x0d9278500d927850),
    U64(0x9bcc5f6a9bcc5f6a), U64(0x62467e5462467e54),
    U64(0xc2138df6c2138df6), U64(0xe8b8d890e8b8d890),
    U64(0x5ef7392e5ef7392e), U64(0xf5afc382f5afc382),
    U64(0xbe805d9fbe805d9f), U64(0x7c93d0697c93d069),
    U64(0xa92dd56fa92dd56f), U64(0xb31225cfb31225cf),
    U64(0x3b99acc83b99acc8), U64(0xa77d1810a77d1810),
    U64(0x6e639ce86e639ce8), U64(0x7bbb3bdb7bbb3bdb),
    U64(0x097826cd097826cd), U64(0xf418596ef418596e),
    U64(0x01b79aec01b79aec), U64(0xa89a4f83a89a4f83),
    U64(0x656e95e6656e95e6), U64(0x7ee6ffaa7ee6ffaa),
    U64(0x08cfbc2108cfbc21), U64(0xe6e815efe6e815ef),
    U64(0xd99be7bad99be7ba), U64(0xce366f4ace366f4a),
    U64(0xd4099fead4099fea), U64(0xd67cb029d67cb029),
    U64(0xafb2a431afb2a431), U64(0x31233f2a31233f2a),
    U64(0x3094a5c63094a5c6), U64(0xc066a235c066a235),
    U64(0x37bc4e7437bc4e74), U64(0xa6ca82fca6ca82fc),
    U64(0xb0d090e0b0d090e0), U64(0x15d8a73315d8a733),
    U64(0x4a9804f14a9804f1), U64(0xf7daec41f7daec41),
    U64(0x0e50cd7f0e50cd7f), U64(0x2ff691172ff69117),
    U64(0x8dd64d768dd64d76), U64(0x4db0ef434db0ef43),
    U64(0x544daacc544daacc), U64(0xdf0496e4df0496e4),
    U64(0xe3b5d19ee3b5d19e), U64(0x1b886a4c1b886a4c),
    U64(0xb81f2cc1b81f2cc1), U64(0x7f5165467f516546),
    U64(0x04ea5e9d04ea5e9d), U64(0x5d358c015d358c01),
    U64(0x737487fa737487fa), U64(0x2e410bfb2e410bfb),
    U64(0x5a1d67b35a1d67b3), U64(0x52d2db9252d2db92),
    U64(0x335610e9335610e9), U64(0x1347d66d1347d66d),
    U64(0x8c61d79a8c61d79a), U64(0x7a0ca1377a0ca137),
    U64(0x8e14f8598e14f859), U64(0x893c13eb893c13eb),
    U64(0xee27a9ceee27a9ce), U64(0x35c961b735c961b7),
    U64(0xede51ce1ede51ce1), U64(0x3cb1477a3cb1477a),
    U64(0x59dfd29c59dfd29c), U64(0x3f73f2553f73f255),
    U64(0x79ce141879ce1418), U64(0xbf37c773bf37c773),
    U64(0xeacdf753eacdf753), U64(0x5baafd5f5baafd5f),
    U64(0x146f3ddf146f3ddf), U64(0x86db447886db4478),
    U64(0x81f3afca81f3afca), U64(0x3ec468b93ec468b9),
    U64(0x2c3424382c342438), U64(0x5f40a3c25f40a3c2),
    U64(0x72c31d1672c31d16), U64(0x0c25e2bc0c25e2bc),
    U64(0x8b493c288b493c28), U64(0x41950dff41950dff),
    U64(0x7101a8397101a839), U64(0xdeb30c08deb30c08),
    U64(0x9ce4b4d89ce4b4d8), U64(0x90c1566490c15664),
    U64(0x6184cb7b6184cb7b), U64(0x70b632d570b632d5),
    U64(0x745c6c48745c6c48), U64(0x4257b8d04257b8d0)
};
static const u8 Td4[256] = {
    0x52U, 0x09U, 0x6aU, 0xd5U, 0x30U, 0x36U, 0xa5U, 0x38U,
    0xbfU, 0x40U, 0xa3U, 0x9eU, 0x81U, 0xf3U, 0xd7U, 0xfbU,
    0x7cU, 0xe3U, 0x39U, 0x82U, 0x9bU, 0x2fU, 0xffU, 0x87U,
    0x34U, 0x8eU, 0x43U, 0x44U, 0xc4U, 0xdeU, 0xe9U, 0xcbU,
    0x54U, 0x7bU, 0x94U, 0x32U, 0xa6U, 0xc2U, 0x23U, 0x3dU,
    0xeeU, 0x4cU, 0x95U, 0x0bU, 0x42U, 0xfaU, 0xc3U, 0x4eU,
    0x08U, 0x2eU, 0xa1U, 0x66U, 0x28U, 0xd9U, 0x24U, 0xb2U,
    0x76U, 0x5bU, 0xa2U, 0x49U, 0x6dU, 0x8bU, 0xd1U, 0x25U,
    0x72U, 0xf8U, 0xf6U, 0x64U, 0x86U, 0x68U, 0x98U, 0x16U,
    0xd4U, 0xa4U, 0x5cU, 0xccU, 0x5dU, 0x65U, 0xb6U, 0x92U,
    0x6cU, 0x70U, 0x48U, 0x50U, 0xfdU, 0xedU, 0xb9U, 0xdaU,
    0x5eU, 0x15U, 0x46U, 0x57U, 0xa7U, 0x8dU, 0x9dU, 0x84U,
    0x90U, 0xd8U, 0xabU, 0x00U, 0x8cU, 0xbcU, 0xd3U, 0x0aU,
    0xf7U, 0xe4U, 0x58U, 0x05U, 0xb8U, 0xb3U, 0x45U, 0x06U,
    0xd0U, 0x2cU, 0x1eU, 0x8fU, 0xcaU, 0x3fU, 0x0fU, 0x02U,
    0xc1U, 0xafU, 0xbdU, 0x03U, 0x01U, 0x13U, 0x8aU, 0x6bU,
    0x3aU, 0x91U, 0x11U, 0x41U, 0x4fU, 0x67U, 0xdcU, 0xeaU,
    0x97U, 0xf2U, 0xcfU, 0xceU, 0xf0U, 0xb4U, 0xe6U, 0x73U,
    0x96U, 0xacU, 0x74U, 0x22U, 0xe7U, 0xadU, 0x35U, 0x85U,
    0xe2U, 0xf9U, 0x37U, 0xe8U, 0x1cU, 0x75U, 0xdfU, 0x6eU,
    0x47U, 0xf1U, 0x1aU, 0x71U, 0x1dU, 0x29U, 0xc5U, 0x89U,
    0x6fU, 0xb7U, 0x62U, 0x0eU, 0xaaU, 0x18U, 0xbeU, 0x1bU,
    0xfcU, 0x56U, 0x3eU, 0x4bU, 0xc6U, 0xd2U, 0x79U, 0x20U,
    0x9aU, 0xdbU, 0xc0U, 0xfeU, 0x78U, 0xcdU, 0x5aU, 0xf4U,
    0x1fU, 0xddU, 0xa8U, 0x33U, 0x88U, 0x07U, 0xc7U, 0x31U,
    0xb1U, 0x12U, 0x10U, 0x59U, 0x27U, 0x80U, 0xecU, 0x5fU,
    0x60U, 0x51U, 0x7fU, 0xa9U, 0x19U, 0xb5U, 0x4aU, 0x0dU,
    0x2dU, 0xe5U, 0x7aU, 0x9fU, 0x93U, 0xc9U, 0x9cU, 0xefU,
    0xa0U, 0xe0U, 0x3bU, 0x4dU, 0xaeU, 0x2aU, 0xf5U, 0xb0U,
    0xc8U, 0xebU, 0xbbU, 0x3cU, 0x83U, 0x53U, 0x99U, 0x61U,
    0x17U, 0x2bU, 0x04U, 0x7eU, 0xbaU, 0x77U, 0xd6U, 0x26U,
    0xe1U, 0x69U, 0x14U, 0x63U, 0x55U, 0x21U, 0x0cU, 0x7dU
};

static const u32 rcon[] = {
    0x00000001U, 0x00000002U, 0x00000004U, 0x00000008U,
    0x00000010U, 0x00000020U, 0x00000040U, 0x00000080U,
    0x0000001bU, 0x00000036U, /* for 128-bit blocks, Rijndael never uses more than 10 rcon values */
};

/**
 * Expand the cipher key into the encryption key schedule.
 */
int AES_set_encrypt_key(const unsigned char *userKey, const int bits,
			AES_KEY *key) {

	u32 *rk;
   	int i = 0;
	u32 temp;

	if (!userKey || !key)
		return -1;
	if (bits != 128 && bits != 192 && bits != 256)
		return -2;

	rk = key->rd_key;

	if (bits==128)
		key->rounds = 10;
	else if (bits==192)
		key->rounds = 12;
	else
		key->rounds = 14;

	rk[0] = GETU32(userKey     );
	rk[1] = GETU32(userKey +  4);
	rk[2] = GETU32(userKey +  8);
	rk[3] = GETU32(userKey + 12);
	if (bits == 128) {
		while (1) {
			temp  = rk[3];
			rk[4] = rk[0] ^
				(Te4[(temp >>  8) & 0xff]      ) ^
				(Te4[(temp >> 16) & 0xff] <<  8) ^
				(Te4[(temp >> 24)       ] << 16) ^
				(Te4[(temp      ) & 0xff] << 24) ^
				rcon[i];
			rk[5] = rk[1] ^ rk[4];
			rk[6] = rk[2] ^ rk[5];
			rk[7] = rk[3] ^ rk[6];
			if (++i == 10) {
				return 0;
			}
			rk += 4;
		}
	}
	rk[4] = GETU32(userKey + 16);
	rk[5] = GETU32(userKey + 20);
	if (bits == 192) {
		while (1) {
			temp = rk[ 5];
			rk[ 6] = rk[ 0] ^
				(Te4[(temp >>  8) & 0xff]      ) ^
				(Te4[(temp >> 16) & 0xff] <<  8) ^
				(Te4[(temp >> 24)       ] << 16) ^
				(Te4[(temp      ) & 0xff] << 24) ^
				rcon[i];
			rk[ 7] = rk[ 1] ^ rk[ 6];
			rk[ 8] = rk[ 2] ^ rk[ 7];
			rk[ 9] = rk[ 3] ^ rk[ 8];
			if (++i == 8) {
				return 0;
			}
			rk[10] = rk[ 4] ^ rk[ 9];
			rk[11] = rk[ 5] ^ rk[10];
			rk += 6;
		}
	}
	rk[6] = GETU32(userKey + 24);
	rk[7] = GETU32(userKey + 28);
	if (bits == 256) {
		while (1) {
			temp = rk[ 7];
			rk[ 8] = rk[ 0] ^
				(Te4[(temp >>  8) & 0xff]      ) ^
				(Te4[(temp >> 16) & 0xff] <<  8) ^
				(Te4[(temp >> 24)       ] << 16) ^
				(Te4[(temp      ) & 0xff] << 24) ^
				rcon[i];
			rk[ 9] = rk[ 1] ^ rk[ 8];
			rk[10] = rk[ 2] ^ rk[ 9];
			rk[11] = rk[ 3] ^ rk[10];
			if (++i == 7) {
				return 0;
			}
			temp = rk[11];
			rk[12] = rk[ 4] ^
				(Te4[(temp      ) & 0xff]      ) ^
				(Te4[(temp >>  8) & 0xff] <<  8) ^
				(Te4[(temp >> 16) & 0xff] << 16) ^
				(Te4[(temp >> 24)       ] << 24);
			rk[13] = rk[ 5] ^ rk[12];
			rk[14] = rk[ 6] ^ rk[13];
			rk[15] = rk[ 7] ^ rk[14];

			rk += 8;
        	}
	}
	return 0;
}

/**
 * Expand the cipher key into the decryption key schedule.
 */
int AES_set_decrypt_key(const unsigned char *userKey, const int bits,
			 AES_KEY *key) {

        u32 *rk;
	int i, j, status;
	u32 temp;

	/* first, start with an encryption schedule */
	status = AES_set_encrypt_key(userKey, bits, key);
	if (status < 0)
		return status;

	rk = key->rd_key;

	/* invert the order of the round keys: */
	for (i = 0, j = 4*(key->rounds); i < j; i += 4, j -= 4) {
		temp = rk[i    ]; rk[i    ] = rk[j    ]; rk[j    ] = temp;
		temp = rk[i + 1]; rk[i + 1] = rk[j + 1]; rk[j + 1] = temp;
		temp = rk[i + 2]; rk[i + 2] = rk[j + 2]; rk[j + 2] = temp;
		temp = rk[i + 3]; rk[i + 3] = rk[j + 3]; rk[j + 3] = temp;
	}
	/* apply the inverse MixColumn transform to all round keys but the first and the last: */
	for (i = 1; i < (key->rounds); i++) {
		rk += 4;
#if 1
		for (j = 0; j < 4; j++) {
			u32 tp1, tp2, tp4, tp8, tp9, tpb, tpd, tpe, m;

			tp1 = rk[j];
			m = tp1 & 0x80808080;
			tp2 = ((tp1 & 0x7f7f7f7f) << 1) ^
				((m - (m >> 7)) & 0x1b1b1b1b);
			m = tp2 & 0x80808080;
			tp4 = ((tp2 & 0x7f7f7f7f) << 1) ^
				((m - (m >> 7)) & 0x1b1b1b1b);
			m = tp4 & 0x80808080;
			tp8 = ((tp4 & 0x7f7f7f7f) << 1) ^
				((m - (m >> 7)) & 0x1b1b1b1b);
			tp9 = tp8 ^ tp1;
			tpb = tp9 ^ tp2;
			tpd = tp9 ^ tp4;
			tpe = tp8 ^ tp4 ^ tp2;
#if defined(ROTATE)
			rk[j] = tpe ^ ROTATE(tpd,16) ^
				ROTATE(tp9,8) ^ ROTATE(tpb,24);
#else
			rk[j] = tpe ^ (tpd >> 16) ^ (tpd << 16) ^ 
				(tp9 >> 24) ^ (tp9 << 8) ^
				(tpb >> 8) ^ (tpb << 24);
#endif
		}
#else
		rk[0] =
			Td0[Te2[(rk[0]      ) & 0xff] & 0xff] ^
			Td1[Te2[(rk[0] >>  8) & 0xff] & 0xff] ^
			Td2[Te2[(rk[0] >> 16) & 0xff] & 0xff] ^
			Td3[Te2[(rk[0] >> 24)       ] & 0xff];
		rk[1] =
			Td0[Te2[(rk[1]      ) & 0xff] & 0xff] ^
			Td1[Te2[(rk[1] >>  8) & 0xff] & 0xff] ^
			Td2[Te2[(rk[1] >> 16) & 0xff] & 0xff] ^
			Td3[Te2[(rk[1] >> 24)       ] & 0xff];
		rk[2] =
			Td0[Te2[(rk[2]      ) & 0xff] & 0xff] ^
			Td1[Te2[(rk[2] >>  8) & 0xff] & 0xff] ^
			Td2[Te2[(rk[2] >> 16) & 0xff] & 0xff] ^
			Td3[Te2[(rk[2] >> 24)       ] & 0xff];
		rk[3] =
			Td0[Te2[(rk[3]      ) & 0xff] & 0xff] ^
			Td1[Te2[(rk[3] >>  8) & 0xff] & 0xff] ^
			Td2[Te2[(rk[3] >> 16) & 0xff] & 0xff] ^
			Td3[Te2[(rk[3] >> 24)       ] & 0xff];
#endif
	}
	return 0;
}

/*
 * Encrypt a single block
 * in and out can overlap
 */
void AES_encrypt(const unsigned char *in, unsigned char *out,
		 const AES_KEY *key) {

	const u32 *rk;
	u32 s0, s1, s2, s3, t[4];
	int r;

	assert(in && out && key);
	rk = key->rd_key;

	/*
	 * map byte array block to cipher state
	 * and add initial round key:
	 */
	s0 = GETU32(in     ) ^ rk[0];
	s1 = GETU32(in +  4) ^ rk[1];
	s2 = GETU32(in +  8) ^ rk[2];
	s3 = GETU32(in + 12) ^ rk[3];

#if defined(AES_COMPACT_IN_OUTER_ROUNDS)
	prefetch256(Te4);

	t[0] =	Te4[(s0      ) & 0xff]       ^
		Te4[(s1 >>  8) & 0xff] <<  8 ^
		Te4[(s2 >> 16) & 0xff] << 16 ^
		Te4[(s3 >> 24)       ] << 24;
	t[1] =	Te4[(s1      ) & 0xff]       ^
		Te4[(s2 >>  8) & 0xff] <<  8 ^
		Te4[(s3 >> 16) & 0xff] << 16 ^
		Te4[(s0 >> 24)       ] << 24;
	t[2] =	Te4[(s2      ) & 0xff]       ^
		Te4[(s3 >>  8) & 0xff] <<  8 ^
		Te4[(s0 >> 16) & 0xff] << 16 ^
		Te4[(s1 >> 24)       ] << 24;
	t[3] =	Te4[(s3      ) & 0xff]       ^
		Te4[(s0 >>  8) & 0xff] <<  8 ^
		Te4[(s1 >> 16) & 0xff] << 16 ^
		Te4[(s2 >> 24)       ] << 24;

	/* now do the linear transform using words */
	{	int i;
		u32 r0, r1, r2;

		for (i = 0; i < 4; i++) {
			r0 = t[i];
			r1 = r0 & 0x80808080;
			r2 = ((r0 & 0x7f7f7f7f) << 1) ^
				((r1 - (r1 >> 7)) & 0x1b1b1b1b);
#if defined(ROTATE)
			t[i] = r2 ^ ROTATE(r2,24) ^ ROTATE(r0,24) ^
				ROTATE(r0,16) ^ ROTATE(r0,8);
#else
			t[i] = r2 ^ ((r2 ^ r0) << 24) ^ ((r2 ^ r0) >> 8) ^
				(r0 << 16) ^ (r0 >> 16) ^
				(r0 << 8) ^ (r0 >> 24);
#endif
			t[i] ^= rk[4+i];
		}
	}