tomcrypt/aes.c

/* AES implementation by Tom St Denis 
 *
 * Derived from the Public Domain source code by
 
---  
  * rijndael-alg-fst.c
  *
  * @version 3.0 (December 2000)
  *
  * Optimised ANSI C code for the Rijndael cipher (now AES)
  *
  * @author Vincent Rijmen <vincent.rijmen@esat.kuleuven.ac.be>
  * @author Antoon Bosselaers <antoon.bosselaers@esat.kuleuven.ac.be>
  * @author Paulo Barreto <paulo.barreto@terra.com.br>
---
 */

#include "mycrypt.h"

#ifdef RIJNDAEL

const struct _cipher_descriptor rijndael_desc =
{
    "rijndael",
    6,
    16, 32, 16, 10,
    &rijndael_setup,
    &rijndael_ecb_encrypt,
    &rijndael_ecb_decrypt,
    &rijndael_test,
    &rijndael_keysize
};

const struct _cipher_descriptor aes_desc =
{
    "aes",
    6,
    16, 32, 16, 10,
    &rijndael_setup,
    &rijndael_ecb_encrypt,
    &rijndael_ecb_decrypt,
    &rijndael_test,
    &rijndael_keysize
};

#include "aes_tab.c"

int rijndael_setup(const unsigned char *key, int keylen, int rounds, symmetric_key *skey)
{
    int i = 0, j;
    unsigned long temp, *rk;
    
    _ARGCHK(key != NULL);
    _ARGCHK(skey != NULL);
    
    if (keylen != 16 && keylen != 24 && keylen != 32) {
       return CRYPT_INVALID_KEYSIZE;
    }
    
    if (rounds != 0 && rounds != (10 + ((keylen/8)-2)*2)) {
       return CRYPT_INVALID_ROUNDS;
    }
    
    skey->rijndael.Nr = 10 + ((keylen/8)-2)*2;
        
    /* setup the forward key */
    rk                = skey->rijndael.eK;
    LOAD32H(rk[0], key     );
    LOAD32H(rk[1], key +  4);
    LOAD32H(rk[2], key +  8);
    LOAD32H(rk[3], key + 12);
    if (keylen == 16) {
        for (;;) {
            temp  = rk[3];
            rk[4] = rk[0] ^
                (Te4[(temp >> 16) & 0xff] & 0xff000000) ^
                (Te4[(temp >>  8) & 0xff] & 0x00ff0000) ^
                (Te4[(temp      ) & 0xff] & 0x0000ff00) ^
                (Te4[(temp >> 24)       ] & 0x000000ff) ^
                rcon[i];
            rk[5] = rk[1] ^ rk[4];
            rk[6] = rk[2] ^ rk[5];
            rk[7] = rk[3] ^ rk[6];
            if (++i == 10) {
                break;
            }
            rk += 4;
        }
    } else if (keylen == 24) {
        LOAD32H(rk[4], key + 16);
        LOAD32H(rk[5], key + 20);
        for (;;) {
        #ifdef _MSC_VER
            temp = skey->rijndael.eK[rk - skey->rijndael.eK + 5]; 
        #else
            temp = rk[5];
        #endif
            rk[ 6] = rk[ 0] ^
                (Te4[(temp >> 16) & 0xff] & 0xff000000) ^
                (Te4[(temp >>  8) & 0xff] & 0x00ff0000) ^
                (Te4[(temp      ) & 0xff] & 0x0000ff00) ^
                (Te4[(temp >> 24)       ] & 0x000000ff) ^
                rcon[i];
            rk[ 7] = rk[ 1] ^ rk[ 6];
            rk[ 8] = rk[ 2] ^ rk[ 7];
            rk[ 9] = rk[ 3] ^ rk[ 8];
            if (++i == 8) {
                break;
            }
            rk[10] = rk[ 4] ^ rk[ 9];
            rk[11] = rk[ 5] ^ rk[10];
            rk += 6;
        }
    } else if (keylen == 32) {
        LOAD32H(rk[4], key + 16);
        LOAD32H(rk[5], key + 20);
        LOAD32H(rk[6], key + 24);
        LOAD32H(rk[7], key + 28);
        for (;;) {
        #ifdef _MSC_VER
            temp = skey->rijndael.eK[rk - skey->rijndael.eK + 7]; 
        #else
            temp = rk[7];
        #endif
            rk[ 8] = rk[ 0] ^
                (Te4[(temp >> 16) & 0xff] & 0xff000000) ^
                (Te4[(temp >>  8) & 0xff] & 0x00ff0000) ^
                (Te4[(temp      ) & 0xff] & 0x0000ff00) ^
                (Te4[(temp >> 24)       ] & 0x000000ff) ^
                rcon[i];
            rk[ 9] = rk[ 1] ^ rk[ 8];
            rk[10] = rk[ 2] ^ rk[ 9];
            rk[11] = rk[ 3] ^ rk[10];
            if (++i == 7) {
                break;
            }
            temp = rk[11];
            rk[12] = rk[ 4] ^
                (Te4[(temp >> 24)       ] & 0xff000000) ^
                (Te4[(temp >> 16) & 0xff] & 0x00ff0000) ^
                (Te4[(temp >>  8) & 0xff] & 0x0000ff00) ^
                (Te4[(temp      ) & 0xff] & 0x000000ff);
            rk[13] = rk[ 5] ^ rk[12];
            rk[14] = rk[ 6] ^ rk[13];
            rk[15] = rk[ 7] ^ rk[14];
            rk += 8;
        }
    }
    
    /* setup the inverse key now */
    memcpy(skey->rijndael.dK, skey->rijndael.eK, sizeof(skey->rijndael.eK));
    rk = skey->rijndael.dK;
    
    for (i = 0, j = 4*skey->rijndael.Nr; 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 < skey->rijndael.Nr; i++) {
        rk += 4;
        rk[0] =
            Td0[Te4[(rk[0] >> 24)       ] & 0xff] ^
            Td1[Te4[(rk[0] >> 16) & 0xff] & 0xff] ^
            Td2[Te4[(rk[0] >>  8) & 0xff] & 0xff] ^
            Td3[Te4[(rk[0]      ) & 0xff] & 0xff];
        rk[1] =
            Td0[Te4[(rk[1] >> 24)       ] & 0xff] ^
            Td1[Te4[(rk[1] >> 16) & 0xff] & 0xff] ^
            Td2[Te4[(rk[1] >>  8) & 0xff] & 0xff] ^
            Td3[Te4[(rk[1]      ) & 0xff] & 0xff];
        rk[2] =
            Td0[Te4[(rk[2] >> 24)       ] & 0xff] ^
            Td1[Te4[(rk[2] >> 16) & 0xff] & 0xff] ^
            Td2[Te4[(rk[2] >>  8) & 0xff] & 0xff] ^
            Td3[Te4[(rk[2]      ) & 0xff] & 0xff];
        rk[3] =
            Td0[Te4[(rk[3] >> 24)       ] & 0xff] ^
            Td1[Te4[(rk[3] >> 16) & 0xff] & 0xff] ^
            Td2[Te4[(rk[3] >>  8) & 0xff] & 0xff] ^
            Td3[Te4[(rk[3]      ) & 0xff] & 0xff];
    }
    
    return CRYPT_OK;   
}

void rijndael_ecb_encrypt(const unsigned char *pt, unsigned char *ct, symmetric_key *skey) 
{
    unsigned long s0, s1, s2, s3, t0, t1, t2, t3, *rk;
    int Nr;
    
#ifdef SMALL_CODE
    int r;
#endif 

    _ARGCHK(pt != NULL);
    _ARGCHK(ct != NULL);
    _ARGCHK(skey != NULL);
    
    Nr = skey->rijndael.Nr;
    rk = skey->rijndael.eK;
    
    /*
     * map byte array block to cipher state
     * and add initial round key:
     */
    LOAD32H(s0, pt      ); s0 ^= rk[0];
    LOAD32H(s1, pt  +  4); s1 ^= rk[1];
    LOAD32H(s2, pt  +  8); s2 ^= rk[2];
    LOAD32H(s3, pt  + 12); s3 ^= rk[3];
#ifndef SMALL_CODE
    /* round 1: */
    t0 = Te0[s0 >> 24] ^ Te1[(s1 >> 16) & 0xff] ^ Te2[(s2 >>  8) & 0xff] ^ Te3[s3 & 0xff] ^ rk[ 4];
    t1 = Te0[s1 >> 24] ^ Te1[(s2 >> 16) & 0xff] ^ Te2[(s3 >>  8) & 0xff] ^ Te3[s0 & 0xff] ^ rk[ 5];
    t2 = Te0[s2 >> 24] ^ Te1[(s3 >> 16) & 0xff] ^ Te2[(s0 >>  8) & 0xff] ^ Te3[s1 & 0xff] ^ rk[ 6];
    t3 = Te0[s3 >> 24] ^ Te1[(s0 >> 16) & 0xff] ^ Te2[(s1 >>  8) & 0xff] ^ Te3[s2 & 0xff] ^ rk[ 7];
    /* round 2: */
    s0 = Te0[t0 >> 24] ^ Te1[(t1 >> 16) & 0xff] ^ Te2[(t2 >>  8) & 0xff] ^ Te3[t3 & 0xff] ^ rk[ 8];
    s1 = Te0[t1 >> 24] ^ Te1[(t2 >> 16) & 0xff] ^ Te2[(t3 >>  8) & 0xff] ^ Te3[t0 & 0xff] ^ rk[ 9];
    s2 = Te0[t2 >> 24] ^ Te1[(t3 >> 16) & 0xff] ^ Te2[(t0 >>  8) & 0xff] ^ Te3[t1 & 0xff] ^ rk[10];
    s3 = Te0[t3 >> 24] ^ Te1[(t0 >> 16) & 0xff] ^ Te2[(t1 >>  8) & 0xff] ^ Te3[t2 & 0xff] ^ rk[11];
    /* round 3: */
    t0 = Te0[s0 >> 24] ^ Te1[(s1 >> 16) & 0xff] ^ Te2[(s2 >>  8) & 0xff] ^ Te3[s3 & 0xff] ^ rk[12];
    t1 = Te0[s1 >> 24] ^ Te1[(s2 >> 16) & 0xff] ^ Te2[(s3 >>  8) & 0xff] ^ Te3[s0 & 0xff] ^ rk[13];
    t2 = Te0[s2 >> 24] ^ Te1[(s3 >> 16) & 0xff] ^ Te2[(s0 >>  8) & 0xff] ^ Te3[s1 & 0xff] ^ rk[14];
    t3 = Te0[s3 >> 24] ^ Te1[(s0 >> 16) & 0xff] ^ Te2[(s1 >>  8) & 0xff] ^ Te3[s2 & 0xff] ^ rk[15];
    /* round 4: */
    s0 = Te0[t0 >> 24] ^ Te1[(t1 >> 16) & 0xff] ^ Te2[(t2 >>  8) & 0xff] ^ Te3[t3 & 0xff] ^ rk[16];
    s1 = Te0[t1 >> 24] ^ Te1[(t2 >> 16) & 0xff] ^ Te2[(t3 >>  8) & 0xff] ^ Te3[t0 & 0xff] ^ rk[17];
    s2 = Te0[t2 >> 24] ^ Te1[(t3 >> 16) & 0xff] ^ Te2[(t0 >>  8) & 0xff] ^ Te3[t1 & 0xff] ^ rk[18];
    s3 = Te0[t3 >> 24] ^ Te1[(t0 >> 16) & 0xff] ^ Te2[(t1 >>  8) & 0xff] ^ Te3[t2 & 0xff] ^ rk[19];
    /* round 5: */
    t0 = Te0[s0 >> 24] ^ Te1[(s1 >> 16) & 0xff] ^ Te2[(s2 >>  8) & 0xff] ^ Te3[s3 & 0xff] ^ rk[20];
    t1 = Te0[s1 >> 24] ^ Te1[(s2 >> 16) & 0xff] ^ Te2[(s3 >>  8) & 0xff] ^ Te3[s0 & 0xff] ^ rk[21];
    t2 = Te0[s2 >> 24] ^ Te1[(s3 >> 16) & 0xff] ^ Te2[(s0 >>  8) & 0xff] ^ Te3[s1 & 0xff] ^ rk[22];
    t3 = Te0[s3 >> 24] ^ Te1[(s0 >> 16) & 0xff] ^ Te2[(s1 >>  8) & 0xff] ^ Te3[s2 & 0xff] ^ rk[23];
    /* round 6: */
    s0 = Te0[t0 >> 24] ^ Te1[(t1 >> 16) & 0xff] ^ Te2[(t2 >>  8) & 0xff] ^ Te3[t3 & 0xff] ^ rk[24];
    s1 = Te0[t1 >> 24] ^ Te1[(t2 >> 16) & 0xff] ^ Te2[(t3 >>  8) & 0xff] ^ Te3[t0 & 0xff] ^ rk[25];
    s2 = Te0[t2 >> 24] ^ Te1[(t3 >> 16) & 0xff] ^ Te2[(t0 >>  8) & 0xff] ^ Te3[t1 & 0xff] ^ rk[26];
    s3 = Te0[t3 >> 24] ^ Te1[(t0 >> 16) & 0xff] ^ Te2[(t1 >>  8) & 0xff] ^ Te3[t2 & 0xff] ^ rk[27];
    /* round 7: */
    t0 = Te0[s0 >> 24] ^ Te1[(s1 >> 16) & 0xff] ^ Te2[(s2 >>  8) & 0xff] ^ Te3[s3 & 0xff] ^ rk[28];
    t1 = Te0[s1 >> 24] ^ Te1[(s2 >> 16) & 0xff] ^ Te2[(s3 >>  8) & 0xff] ^ Te3[s0 & 0xff] ^ rk[29];
    t2 = Te0[s2 >> 24] ^ Te1[(s3 >> 16) & 0xff] ^ Te2[(s0 >>  8) & 0xff] ^ Te3[s1 & 0xff] ^ rk[30];
    t3 = Te0[s3 >> 24] ^ Te1[(s0 >> 16) & 0xff] ^ Te2[(s1 >>  8) & 0xff] ^ Te3[s2 & 0xff] ^ rk[31];
    /* round 8: */
    s0 = Te0[t0 >> 24] ^ Te1[(t1 >> 16) & 0xff] ^ Te2[(t2 >>  8) & 0xff] ^ Te3[t3 & 0xff] ^ rk[32];
    s1 = Te0[t1 >> 24] ^ Te1[(t2 >> 16) & 0xff] ^ Te2[(t3 >>  8) & 0xff] ^ Te3[t0 & 0xff] ^ rk[33];
    s2 = Te0[t2 >> 24] ^ Te1[(t3 >> 16) & 0xff] ^ Te2[(t0 >>  8) & 0xff] ^ Te3[t1 & 0xff] ^ rk[34];
    s3 = Te0[t3 >> 24] ^ Te1[(t0 >> 16) & 0xff] ^ Te2[(t1 >>  8) & 0xff] ^ Te3[t2 & 0xff] ^ rk[35];
    /* round 9: */
    t0 = Te0[s0 >> 24] ^ Te1[(s1 >> 16) & 0xff] ^ Te2[(s2 >>  8) & 0xff] ^ Te3[s3 & 0xff] ^ rk[36];
    t1 = Te0[s1 >> 24] ^ Te1[(s2 >> 16) & 0xff] ^ Te2[(s3 >>  8) & 0xff] ^ Te3[s0 & 0xff] ^ rk[37];
    t2 = Te0[s2 >> 24] ^ Te1[(s3 >> 16) & 0xff] ^ Te2[(s0 >>  8) & 0xff] ^ Te3[s1 & 0xff] ^ rk[38];
    t3 = Te0[s3 >> 24] ^ Te1[(s0 >> 16) & 0xff] ^ Te2[(s1 >>  8) & 0xff] ^ Te3[s2 & 0xff] ^ rk[39];
    if (Nr > 10) {
        /* round 10: */
        s0 = Te0[t0 >> 24] ^ Te1[(t1 >> 16) & 0xff] ^ Te2[(t2 >>  8) & 0xff] ^ Te3[t3 & 0xff] ^ rk[40];
        s1 = Te0[t1 >> 24] ^ Te1[(t2 >> 16) & 0xff] ^ Te2[(t3 >>  8) & 0xff] ^ Te3[t0 & 0xff] ^ rk[41];
        s2 = Te0[t2 >> 24] ^ Te1[(t3 >> 16) & 0xff] ^ Te2[(t0 >>  8) & 0xff] ^ Te3[t1 & 0xff] ^ rk[42];
        s3 = Te0[t3 >> 24] ^ Te1[(t0 >> 16) & 0xff] ^ Te2[(t1 >>  8) & 0xff] ^ Te3[t2 & 0xff] ^ rk[43];
        /* round 11: */
        t0 = Te0[s0 >> 24] ^ Te1[(s1 >> 16) & 0xff] ^ Te2[(s2 >>  8) & 0xff] ^ Te3[s3 & 0xff] ^ rk[44];
        t1 = Te0[s1 >> 24] ^ Te1[(s2 >> 16) & 0xff] ^ Te2[(s3 >>  8) & 0xff] ^ Te3[s0 & 0xff] ^ rk[45];
        t2 = Te0[s2 >> 24] ^ Te1[(s3 >> 16) & 0xff] ^ Te2[(s0 >>  8) & 0xff] ^ Te3[s1 & 0xff] ^ rk[46];
        t3 = Te0[s3 >> 24] ^ Te1[(s0 >> 16) & 0xff] ^ Te2[(s1 >>  8) & 0xff] ^ Te3[s2 & 0xff] ^ rk[47];
        if (Nr > 12) {
            /* round 12: */
            s0 = Te0[t0 >> 24] ^ Te1[(t1 >> 16) & 0xff] ^ Te2[(t2 >>  8) & 0xff] ^ Te3[t3 & 0xff] ^ rk[48];
            s1 = Te0[t1 >> 24] ^ Te1[(t2 >> 16) & 0xff] ^ Te2[(t3 >>  8) & 0xff] ^ Te3[t0 & 0xff] ^ rk[49];
            s2 = Te0[t2 >> 24] ^ Te1[(t3 >> 16) & 0xff] ^ Te2[(t0 >>  8) & 0xff] ^ Te3[t1 & 0xff] ^ rk[50];
            s3 = Te0[t3 >> 24] ^ Te1[(t0 >> 16) & 0xff] ^ Te2[(t1 >>  8) & 0xff] ^ Te3[t2 & 0xff] ^ rk[51];
            /* round 13: */
            t0 = Te0[s0 >> 24] ^ Te1[(s1 >> 16) & 0xff] ^ Te2[(s2 >>  8) & 0xff] ^ Te3[s3 & 0xff] ^ rk[52];
            t1 = Te0[s1 >> 24] ^ Te1[(s2 >> 16) & 0xff] ^ Te2[(s3 >>  8) & 0xff] ^ Te3[s0 & 0xff] ^ rk[53];
            t2 = Te0[s2 >> 24] ^ Te1[(s3 >> 16) & 0xff] ^ Te2[(s0 >>  8) & 0xff] ^ Te3[s1 & 0xff] ^ rk[54];
            t3 = Te0[s3 >> 24] ^ Te1[(s0 >> 16) & 0xff] ^ Te2[(s1 >>  8) & 0xff] ^ Te3[s2 & 0xff] ^ rk[55];
        }
    }
    rk += Nr << 2;
#else  /* SMALL_CODE */
    /*
     * Nr - 1 full rounds:
     */
    r = Nr >> 1;
    for (;;) {

/* Both of these blocks are equivalent except the top is more friendlier for x86 processors */
#if 1
        t0 = rk[4]; t1 = rk[5]; t2 = rk[6]; t3 = rk[7];
        t1 ^= Te3[(s0      ) & 0xFF]; t2 ^= Te2[(s0 >> 8) & 0xFF]; t3 ^= Te1[(s0 >> 16) & 0xFF]; t0 ^= Te0[(s0 >> 24)];
        t2 ^= Te3[(s1      ) & 0xFF]; t3 ^= Te2[(s1 >> 8) & 0xFF]; t0 ^= Te1[(s1 >> 16) & 0xFF]; t1 ^= Te0[(s1 >> 24)];
        t3 ^= Te3[(s2      ) & 0xFF]; t0 ^= Te2[(s2 >> 8) & 0xFF]; t1 ^= Te1[(s2 >> 16) & 0xFF]; t2 ^= Te0[(s2 >> 24)];
        t0 ^= Te3[(s3      ) & 0xFF]; t1 ^= Te2[(s3 >> 8) & 0xFF]; t2 ^= Te1[(s3 >> 16) & 0xFF]; t3 ^= Te0[(s3 >> 24)];
#else
        t0 =
            Te0[(s0 >> 24)       ] ^
            Te1[(s1 >> 16) & 0xff] ^
            Te2[(s2 >>  8) & 0xff] ^
            Te3[(s3      ) & 0xff] ^
            rk[4];
        t1 =
            Te0[(s1 >> 24)       ] ^
            Te1[(s2 >> 16) & 0xff] ^
            Te2[(s3 >>  8) & 0xff] ^
            Te3[(s0      ) & 0xff] ^
            rk[5];
        t2 =
            Te0[(s2 >> 24)       ] ^
            Te1[(s3 >> 16) & 0xff] ^
            Te2[(s0 >>  8) & 0xff] ^
            Te3[(s1      ) & 0xff] ^
            rk[6];
        t3 =
            Te0[(s3 >> 24)       ] ^
            Te1[(s0 >> 16) & 0xff] ^
            Te2[(s1 >>  8) & 0xff] ^
            Te3[(s2      ) & 0xff] ^
            rk[7];
#endif
       
        rk += 8;
        if (--r == 0) {
            break;
        }
        
/* this second half optimization actually makes it slower on the Athlon, use with caution. */
#if 0
        s1 = rk[1]; s2 = rk[2]; s3 = rk[3]; s0 = rk[0]; 
        s1 ^= Te3[(t0      ) & 0xFF]; s2 ^= Te2[(t0 >> 8) & 0xFF]; s3 ^= Te1[(t0 >> 16) & 0xFF]; s0 ^= Te0[(t0 >> 24)];
        s2 ^= Te3[(t1      ) & 0xFF]; s3 ^= Te2[(t1 >> 8) & 0xFF]; s0 ^= Te1[(t1 >> 16) & 0xFF]; s1 ^= Te0[(t1 >> 24)];
        s3 ^= Te3[(t2      ) & 0xFF]; s0 ^= Te2[(t2 >> 8) & 0xFF]; s1 ^= Te1[(t2 >> 16) & 0xFF]; s2 ^= Te0[(t2 >> 24)];
        s0 ^= Te3[(t3      ) & 0xFF]; s1 ^= Te2[(t3 >> 8) & 0xFF]; s2 ^= Te1[(t3 >> 16) & 0xFF]; s3 ^= Te0[(t3 >> 24)];
#else
        s0 =
            Te0[(t0 >> 24)       ] ^
            Te1[(t1 >> 16) & 0xff] ^
            Te2[(t2 >>  8) & 0xff] ^
            Te3[(t3      ) & 0xff] ^
            rk[0];
        s1 =
            Te0[(t1 >> 24)       ] ^
            Te1[(t2 >> 16) & 0xff] ^
            Te2[(t3 >>  8) & 0xff] ^
            Te3[(t0      ) & 0xff] ^
            rk[1];
        s2 =
            Te0[(t2 >> 24)       ] ^
            Te1[(t3 >> 16) & 0xff] ^
            Te2[(t0 >>  8) & 0xff] ^
            Te3[(t1      ) & 0xff] ^
            rk[2];
        s3 =
            Te0[(t3 >> 24)       ] ^
            Te1[(t0 >> 16) & 0xff] ^
            Te2[(t1 >>  8) & 0xff] ^
            Te3[(t2      ) & 0xff] ^
            rk[3];
#endif            
    }
#endif /* SMALL_CODE */
    /*
     * apply last round and
     * map cipher state to byte array block:
     */
    s0 =
        (Te4[(t0 >> 24)       ] & 0xff000000) ^
        (Te4[(t1 >> 16) & 0xff] & 0x00ff0000) ^
        (Te4[(t2 >>  8) & 0xff] & 0x0000ff00) ^
        (Te4[(t3      ) & 0xff] & 0x000000ff) ^
        rk[0];
    STORE32H(s0, ct);
    s1 =
        (Te4[(t1 >> 24)       ] & 0xff000000) ^
        (Te4[(t2 >> 16) & 0xff] & 0x00ff0000) ^
        (Te4[(t3 >>  8) & 0xff] & 0x0000ff00) ^
        (Te4[(t0      ) & 0xff] & 0x000000ff) ^
        rk[1];
    STORE32H(s1, ct+4);
    s2 =
        (Te4[(t2 >> 24)       ] & 0xff000000) ^
        (Te4[(t3 >> 16) & 0xff] & 0x00ff0000) ^
        (Te4[(t0 >>  8) & 0xff] & 0x0000ff00) ^
        (Te4[(t1      ) & 0xff] & 0x000000ff) ^
        rk[2];
    STORE32H(s2, ct+8);
    s3 =
        (Te4[(t3 >> 24)       ] & 0xff000000) ^
        (Te4[(t0 >> 16) & 0xff] & 0x00ff0000) ^
        (Te4[(t1 >>  8) & 0xff] & 0x0000ff00) ^
        (Te4[(t2      ) & 0xff] & 0x000000ff) ^
        rk[3];
    STORE32H(s3, ct+12);
}

void rijndael_ecb_decrypt(const unsigned char *ct, unsigned char *pt, symmetric_key *skey) {
    unsigned long s0, s1, s2, s3, t0, t1, t2, t3, *rk;
    int Nr;
#ifdef SMALL_CODE
    int r;
#endif /* SMALL_CODE */

    _ARGCHK(pt != NULL);
    _ARGCHK(ct != NULL);
    _ARGCHK(skey != NULL);
    
    Nr = skey->rijndael.Nr;
    rk = skey->rijndael.dK;

    /*
     * map byte array block to cipher state
     * and add initial round key:
     */
    LOAD32H(s0, ct      ); s0 ^= rk[0];
    LOAD32H(s1, ct  +  4); s1 ^= rk[1];
    LOAD32H(s2, ct  +  8); s2 ^= rk[2];
    LOAD32H(s3, ct  + 12); s3 ^= rk[3];
#ifndef SMALL_CODE
    /* round 1: */
    t0 = Td0[s0 >> 24] ^ Td1[(s3 >> 16) & 0xff] ^ Td2[(s2 >>  8) & 0xff] ^ Td3[s1 & 0xff] ^ rk[ 4];
    t1 = Td0[s1 >> 24] ^ Td1[(s0 >> 16) & 0xff] ^ Td2[(s3 >>  8) & 0xff] ^ Td3[s2 & 0xff] ^ rk[ 5];
    t2 = Td0[s2 >> 24] ^ Td1[(s1 >> 16) & 0xff] ^ Td2[(s0 >>  8) & 0xff] ^ Td3[s3 & 0xff] ^ rk[ 6];
    t3 = Td0[s3 >> 24] ^ Td1[(s2 >> 16) & 0xff] ^ Td2[(s1 >>  8) & 0xff] ^ Td3[s0 & 0xff] ^ rk[ 7];
    /* round 2: */
    s0 = Td0[t0 >> 24] ^ Td1[(t3 >> 16) & 0xff] ^ Td2[(t2 >>  8) & 0xff] ^ Td3[t1 & 0xff] ^ rk[ 8];
    s1 = Td0[t1 >> 24] ^ Td1[(t0 >> 16) & 0xff] ^ Td2[(t3 >>  8) & 0xff] ^ Td3[t2 & 0xff] ^ rk[ 9];
    s2 = Td0[t2 >> 24] ^ Td1[(t1 >> 16) & 0xff] ^ Td2[(t0 >>  8) & 0xff] ^ Td3[t3 & 0xff] ^ rk[10];
    s3 = Td0[t3 >> 24] ^ Td1[(t2 >> 16) & 0xff] ^ Td2[(t1 >>  8) & 0xff] ^ Td3[t0 & 0xff] ^ rk[11];
    /* round 3: */
    t0 = Td0[s0 >> 24] ^ Td1[(s3 >> 16) & 0xff] ^ Td2[(s2 >>  8) & 0xff] ^ Td3[s1 & 0xff] ^ rk[12];
    t1 = Td0[s1 >> 24] ^ Td1[(s0 >> 16) & 0xff] ^ Td2[(s3 >>  8) & 0xff] ^ Td3[s2 & 0xff] ^ rk[13];
    t2 = Td0[s2 >> 24] ^ Td1[(s1 >> 16) & 0xff] ^ Td2[(s0 >>  8) & 0xff] ^ Td3[s3 & 0xff] ^ rk[14];
    t3 = Td0[s3 >> 24] ^ Td1[(s2 >> 16) & 0xff] ^ Td2[(s1 >>  8) & 0xff] ^ Td3[s0 & 0xff] ^ rk[15];
    /* round 4: */
    s0 = Td0[t0 >> 24] ^ Td1[(t3 >> 16) & 0xff] ^ Td2[(t2 >>  8) & 0xff] ^ Td3[t1 & 0xff] ^ rk[16];
    s1 = Td0[t1 >> 24] ^ Td1[(t0 >> 16) & 0xff] ^ Td2[(t3 >>  8) & 0xff] ^ Td3[t2 & 0xff] ^ rk[17];
    s2 = Td0[t2 >> 24] ^ Td1[(t1 >> 16) & 0xff] ^ Td2[(t0 >>  8) & 0xff] ^ Td3[t3 & 0xff] ^ rk[18];
    s3 = Td0[t3 >> 24] ^ Td1[(t2 >> 16) & 0xff] ^ Td2[(t1 >>  8) & 0xff] ^ Td3[t0 & 0xff] ^ rk[19];
    /* round 5: */
    t0 = Td0[s0 >> 24] ^ Td1[(s3 >> 16) & 0xff] ^ Td2[(s2 >>  8) & 0xff] ^ Td3[s1 & 0xff] ^ rk[20];
    t1 = Td0[s1 >> 24] ^ Td1[(s0 >> 16) & 0xff] ^ Td2[(s3 >>  8) & 0xff] ^ Td3[s2 & 0xff] ^ rk[21];
    t2 = Td0[s2 >> 24] ^ Td1[(s1 >> 16) & 0xff] ^ Td2[(s0 >>  8) & 0xff] ^ Td3[s3 & 0xff] ^ rk[22];
    t3 = Td0[s3 >> 24] ^ Td1[(s2 >> 16) & 0xff] ^ Td2[(s1 >>  8) & 0xff] ^ Td3[s0 & 0xff] ^ rk[23];
    /* round 6: */
    s0 = Td0[t0 >> 24] ^ Td1[(t3 >> 16) & 0xff] ^ Td2[(t2 >>  8) & 0xff] ^ Td3[t1 & 0xff] ^ rk[24];
    s1 = Td0[t1 >> 24] ^ Td1[(t0 >> 16) & 0xff] ^ Td2[(t3 >>  8) & 0xff] ^ Td3[t2 & 0xff] ^ rk[25];
    s2 = Td0[t2 >> 24] ^ Td1[(t1 >> 16) & 0xff] ^ Td2[(t0 >>  8) & 0xff] ^ Td3[t3 & 0xff] ^ rk[26];
    s3 = Td0[t3 >> 24] ^ Td1[(t2 >> 16) & 0xff] ^ Td2[(t1 >>  8) & 0xff] ^ Td3[t0 & 0xff] ^ rk[27];
    /* round 7: */
    t0 = Td0[s0 >> 24] ^ Td1[(s3 >> 16) & 0xff] ^ Td2[(s2 >>  8) & 0xff] ^ Td3[s1 & 0xff] ^ rk[28];
    t1 = Td0[s1 >> 24] ^ Td1[(s0 >> 16) & 0xff] ^ Td2[(s3 >>  8) & 0xff] ^ Td3[s2 & 0xff] ^ rk[29];
    t2 = Td0[s2 >> 24] ^ Td1[(s1 >> 16) & 0xff] ^ Td2[(s0 >>  8) & 0xff] ^ Td3[s3 & 0xff] ^ rk[30];
    t3 = Td0[s3 >> 24] ^ Td1[(s2 >> 16) & 0xff] ^ Td2[(s1 >>  8) & 0xff] ^ Td3[s0 & 0xff] ^ rk[31];
    /* round 8: */
    s0 = Td0[t0 >> 24] ^ Td1[(t3 >> 16) & 0xff] ^ Td2[(t2 >>  8) & 0xff] ^ Td3[t1 & 0xff] ^ rk[32];
    s1 = Td0[t1 >> 24] ^ Td1[(t0 >> 16) & 0xff] ^ Td2[(t3 >>  8) & 0xff] ^ Td3[t2 & 0xff] ^ rk[33];
    s2 = Td0[t2 >> 24] ^ Td1[(t1 >> 16) & 0xff] ^ Td2[(t0 >>  8) & 0xff] ^ Td3[t3 & 0xff] ^ rk[34];
    s3 = Td0[t3 >> 24] ^ Td1[(t2 >> 16) & 0xff] ^ Td2[(t1 >>  8) & 0xff] ^ Td3[t0 & 0xff] ^ rk[35];
    /* round 9: */
    t0 = Td0[s0 >> 24] ^ Td1[(s3 >> 16) & 0xff] ^ Td2[(s2 >>  8) & 0xff] ^ Td3[s1 & 0xff] ^ rk[36];
    t1 = Td0[s1 >> 24] ^ Td1[(s0 >> 16) & 0xff] ^ Td2[(s3 >>  8) & 0xff] ^ Td3[s2 & 0xff] ^ rk[37];
    t2 = Td0[s2 >> 24] ^ Td1[(s1 >> 16) & 0xff] ^ Td2[(s0 >>  8) & 0xff] ^ Td3[s3 & 0xff] ^ rk[38];
    t3 = Td0[s3 >> 24] ^ Td1[(s2 >> 16) & 0xff] ^ Td2[(s1 >>  8) & 0xff] ^ Td3[s0 & 0xff] ^ rk[39];
    if (Nr > 10) {
        /* round 10: */
        s0 = Td0[t0 >> 24] ^ Td1[(t3 >> 16) & 0xff] ^ Td2[(t2 >>  8) & 0xff] ^ Td3[t1 & 0xff] ^ rk[40];
        s1 = Td0[t1 >> 24] ^ Td1[(t0 >> 16) & 0xff] ^ Td2[(t3 >>  8) & 0xff] ^ Td3[t2 & 0xff] ^ rk[41];
        s2 = Td0[t2 >> 24] ^ Td1[(t1 >> 16) & 0xff] ^ Td2[(t0 >>  8) & 0xff] ^ Td3[t3 & 0xff] ^ rk[42];
        s3 = Td0[t3 >> 24] ^ Td1[(t2 >> 16) & 0xff] ^ Td2[(t1 >>  8) & 0xff] ^ Td3[t0 & 0xff] ^ rk[43];
        /* round 11: */
        t0 = Td0[s0 >> 24] ^ Td1[(s3 >> 16) & 0xff] ^ Td2[(s2 >>  8) & 0xff] ^ Td3[s1 & 0xff] ^ rk[44];
        t1 = Td0[s1 >> 24] ^ Td1[(s0 >> 16) & 0xff] ^ Td2[(s3 >>  8) & 0xff] ^ Td3[s2 & 0xff] ^ rk[45];
        t2 = Td0[s2 >> 24] ^ Td1[(s1 >> 16) & 0xff] ^ Td2[(s0 >>  8) & 0xff] ^ Td3[s3 & 0xff] ^ rk[46];
        t3 = Td0[s3 >> 24] ^ Td1[(s2 >> 16) & 0xff] ^ Td2[(s1 >>  8) & 0xff] ^ Td3[s0 & 0xff] ^ rk[47];
        if (Nr > 12) {
            /* round 12: */
            s0 = Td0[t0 >> 24] ^ Td1[(t3 >> 16) & 0xff] ^ Td2[(t2 >>  8) & 0xff] ^ Td3[t1 & 0xff] ^ rk[48];
            s1 = Td0[t1 >> 24] ^ Td1[(t0 >> 16) & 0xff] ^ Td2[(t3 >>  8) & 0xff] ^ Td3[t2 & 0xff] ^ rk[49];
            s2 = Td0[t2 >> 24] ^ Td1[(t1 >> 16) & 0xff] ^ Td2[(t0 >>  8) & 0xff] ^ Td3[t3 & 0xff] ^ rk[50];
            s3 = Td0[t3 >> 24] ^ Td1[(t2 >> 16) & 0xff] ^ Td2[(t1 >>  8) & 0xff] ^ Td3[t0 & 0xff] ^ rk[51];
            /* round 13: */
            t0 = Td0[s0 >> 24] ^ Td1[(s3 >> 16) & 0xff] ^ Td2[(s2 >>  8) & 0xff] ^ Td3[s1 & 0xff] ^ rk[52];
            t1 = Td0[s1 >> 24] ^ Td1[(s0 >> 16) & 0xff] ^ Td2[(s3 >>  8) & 0xff] ^ Td3[s2 & 0xff] ^ rk[53];
            t2 = Td0[s2 >> 24] ^ Td1[(s1 >> 16) & 0xff] ^ Td2[(s0 >>  8) & 0xff] ^ Td3[s3 & 0xff] ^ rk[54];
            t3 = Td0[s3 >> 24] ^ Td1[(s2 >> 16) & 0xff] ^ Td2[(s1 >>  8) & 0xff] ^ Td3[s0 & 0xff] ^ rk[55];
        }
    }
    rk += Nr << 2;
#else  /* SMALL_CODE */
    /*
     * Nr - 1 full rounds:
     */
    r = Nr >> 1;
    for (;;) {
        t0 =
            Td0[(s0 >> 24)       ] ^
            Td1[(s3 >> 16) & 0xff] ^
            Td2[(s2 >>  8) & 0xff] ^
            Td3[(s1      ) & 0xff] ^
            rk[4];
        t1 =
            Td0[(s1 >> 24)       ] ^
            Td1[(s0 >> 16) & 0xff] ^
            Td2[(s3 >>  8) & 0xff] ^
            Td3[(s2      ) & 0xff] ^
            rk[5];
        t2 =
            Td0[(s2 >> 24)       ] ^
            Td1[(s1 >> 16) & 0xff] ^
            Td2[(s0 >>  8) & 0xff] ^
            Td3[(s3      ) & 0xff] ^
            rk[6];
        t3 =
            Td0[(s3 >> 24)       ] ^
            Td1[(s2 >> 16) & 0xff] ^
            Td2[(s1 >>  8) & 0xff] ^
            Td3[(s0      ) & 0xff] ^
            rk[7];

        rk += 8;
        if (--r == 0) {
            break;
        }

        s0 =
            Td0[(t0 >> 24)       ] ^
            Td1[(t3 >> 16) & 0xff] ^
            Td2[(t2 >>  8) & 0xff] ^
            Td3[(t1      ) & 0xff] ^
            rk[0];
        s1 =
            Td0[(t1 >> 24)       ] ^
            Td1[(t0 >> 16) & 0xff] ^
            Td2[(t3 >>  8) & 0xff] ^
            Td3[(t2      ) & 0xff] ^
            rk[1];
        s2 =
            Td0[(t2 >> 24)       ] ^
            Td1[(t1 >> 16) & 0xff] ^
            Td2[(t0 >>  8) & 0xff] ^
            Td3[(t3      ) & 0xff] ^
            rk[2];
        s3 =
            Td0[(t3 >> 24)       ] ^
            Td1[(t2 >> 16) & 0xff] ^
            Td2[(t1 >>  8) & 0xff] ^
            Td3[(t0      ) & 0xff] ^
            rk[3];
    }
#endif /* SMALL_CODE */
    /*
     * apply last round and
     * map cipher state to byte array block:
     */
    s0 =
        (Td4[(t0 >> 24)       ] & 0xff000000) ^
        (Td4[(t3 >> 16) & 0xff] & 0x00ff0000) ^
        (Td4[(t2 >>  8) & 0xff] & 0x0000ff00) ^
        (Td4[(t1      ) & 0xff] & 0x000000ff) ^
        rk[0];
    STORE32H(s0, pt);
    s1 =
        (Td4[(t1 >> 24)       ] & 0xff000000) ^
        (Td4[(t0 >> 16) & 0xff] & 0x00ff0000) ^
        (Td4[(t3 >>  8) & 0xff] & 0x0000ff00) ^
        (Td4[(t2      ) & 0xff] & 0x000000ff) ^
        rk[1];
    STORE32H(s1, pt+4);
    s2 =
        (Td4[(t2 >> 24)       ] & 0xff000000) ^
        (Td4[(t1 >> 16) & 0xff] & 0x00ff0000) ^
        (Td4[(t0 >>  8) & 0xff] & 0x0000ff00) ^
        (Td4[(t3      ) & 0xff] & 0x000000ff) ^
        rk[2];
    STORE32H(s2, pt+8);
    s3 =
        (Td4[(t3 >> 24)       ] & 0xff000000) ^
        (Td4[(t2 >> 16) & 0xff] & 0x00ff0000) ^
        (Td4[(t1 >>  8) & 0xff] & 0x0000ff00) ^
        (Td4[(t0      ) & 0xff] & 0x000000ff) ^
        rk[3];
    STORE32H(s3, pt+12);
}

int rijndael_test(void)
{
 #ifndef LTC_TEST
    return CRYPT_NOP;
 #else    
 int err;
 static const struct {
     int keylen;
     unsigned char key[32], pt[16], ct[16];
 } tests[] = {
    { 16,
      { 0x00, 0x01, 0x02, 0x03, 0x04, 0x05, 0x06, 0x07, 
        0x08, 0x09, 0x0a, 0x0b, 0x0c, 0x0d, 0x0e, 0x0f }, 
      { 0x00, 0x11, 0x22, 0x33, 0x44, 0x55, 0x66, 0x77,
        0x88, 0x99, 0xaa, 0xbb, 0xcc, 0xdd, 0xee, 0xff },
      { 0x69, 0xc4, 0xe0, 0xd8, 0x6a, 0x7b, 0x04, 0x30, 
        0xd8, 0xcd, 0xb7, 0x80, 0x70, 0xb4, 0xc5, 0x5a }
    }, { 
      24,
      { 0x00, 0x01, 0x02, 0x03, 0x04, 0x05, 0x06, 0x07, 
        0x08, 0x09, 0x0a, 0x0b, 0x0c, 0x0d, 0x0e, 0x0f,
        0x10, 0x11, 0x12, 0x13, 0x14, 0x15, 0x16, 0x17 },
      { 0x00, 0x11, 0x22, 0x33, 0x44, 0x55, 0x66, 0x77,
        0x88, 0x99, 0xaa, 0xbb, 0xcc, 0xdd, 0xee, 0xff },
      { 0xdd, 0xa9, 0x7c, 0xa4, 0x86, 0x4c, 0xdf, 0xe0, 
        0x6e, 0xaf, 0x70, 0xa0, 0xec, 0x0d, 0x71, 0x91 }
    }, {
      32,
      { 0x00, 0x01, 0x02, 0x03, 0x04, 0x05, 0x06, 0x07, 
        0x08, 0x09, 0x0a, 0x0b, 0x0c, 0x0d, 0x0e, 0x0f,
        0x10, 0x11, 0x12, 0x13, 0x14, 0x15, 0x16, 0x17, 
        0x18, 0x19, 0x1a, 0x1b, 0x1c, 0x1d, 0x1e, 0x1f },
      { 0x00, 0x11, 0x22, 0x33, 0x44, 0x55, 0x66, 0x77,
        0x88, 0x99, 0xaa, 0xbb, 0xcc, 0xdd, 0xee, 0xff },
      { 0x8e, 0xa2, 0xb7, 0xca, 0x51, 0x67, 0x45, 0xbf, 
        0xea, 0xfc, 0x49, 0x90, 0x4b, 0x49, 0x60, 0x89 }
    }
 };
 
 symmetric_key key;
 unsigned char tmp[2][16];
 int i;
 
 for (i = 0; i < (int)(sizeof(tests)/sizeof(tests[0])); i++) {
    zeromem(&key, sizeof(key));
    if ((err = rijndael_setup(tests[i].key, tests[i].keylen, 0, &key)) != CRYPT_OK) { 
       return err;
    }
  
    rijndael_ecb_encrypt(tests[i].pt, tmp[0], &key);
    rijndael_ecb_decrypt(tmp[0], tmp[1], &key);
    if (memcmp(tmp[0], tests[i].ct, 16) || memcmp(tmp[1], tests[i].pt, 16)) { 
#if 0
       printf("\n\nTest %d failed\n", i);
       if (memcmp(tmp[0], tests[i].ct, 16)) {
          printf("CT: ");
          for (i = 0; i < 16; i++) {
             printf("%02x ", tmp[0][i]);
          }
          printf("\n");
       } else {
          printf("PT: ");
          for (i = 0; i < 16; i++) {
             printf("%02x ", tmp[1][i]);
          }
          printf("\n");
       }
#endif       
        return CRYPT_FAIL_TESTVECTOR;
    }
 }       
 return CRYPT_OK;
 #endif
}

int rijndael_keysize(int *desired_keysize)
{
   _ARGCHK(desired_keysize != NULL);

   if (*desired_keysize < 16)
      return CRYPT_INVALID_KEYSIZE;
   if (*desired_keysize < 24) {
      *desired_keysize = 16;
      return CRYPT_OK;
   } else if (*desired_keysize < 32) {
      *desired_keysize = 24;
      return CRYPT_OK;
   } else {
      *desired_keysize = 32;
      return CRYPT_OK;
   }
}

#endif