merge VLIW proofs

Merge branch 'mppa-mul' into mppa-ternary

4 files changed, 32 insertions, 1526 deletions

diff --git a/test/monniaux/bitsliced-aes/bs.c b/test/monniaux/bitsliced-aes/bs.c
index 8df53391..063f36f5 100644
--- a/test/monniaux/bitsliced-aes/bs.c
+++ b/test/monniaux/bitsliced-aes/bs.c
@@ -29,7 +29,6 @@ static inline word_t compcert_ternary(word_t x, word_t v0, word_t v1) {
 #error "endianness not supported"
 #endif
 
-
 void bs_addroundkey(word_t * B, word_t * rk)
 {
     int i;
diff --git a/test/monniaux/bitsliced-aes/one_file/reduce/bitsliced-aes_compute.c b/test/monniaux/bitsliced-aes/one_file/reduce/bitsliced-aes_compute.c
index 44b88d1f..5294ff1d 100644
--- a/test/monniaux/bitsliced-aes/one_file/reduce/bitsliced-aes_compute.c
+++ b/test/monniaux/bitsliced-aes/one_file/reduce/bitsliced-aes_compute.c
@@ -1,1525 +1,32 @@
 #include <stdint.h>
-#include <stdlib.h>
 #include <string.h>
-#include <stdio.h>
-#include "/home/monniaux/work/Kalray/CompCert/test/monniaux/clock.h"
-
-#define EXIT1
-
-void aes_ecb_encrypt(uint8_t * outputb, uint8_t * inputb, size_t size, uint8_t * key);
-void aes_ecb_decrypt(uint8_t * outputb, uint8_t * inputb, size_t size, uint8_t * key);
-
-void aes_ctr_encrypt(uint8_t * outputb, uint8_t * inputb, size_t size, uint8_t * key, uint8_t * iv);
-#define aes_ctr_decrypt(outputb,inputb,size,key,iv) aes_ctr_encrypt(outputb,inputb,size,key,iv)
-
-#define BLOCK_SIZE          128
-#define KEY_SCHEDULE_SIZE   176
-#define WORD_SIZE           64
-#define BS_BLOCK_SIZE       (BLOCK_SIZE * WORD_SIZE / 8)
-#define WORDS_PER_BLOCK     (BLOCK_SIZE / WORD_SIZE)
-
-#if (WORD_SIZE==64)
-    typedef uint64_t    word_t;
-    #define ONE         1ULL
-    #define MUL_SHIFT   6
-    #define WFMT        "lx"
-    #define WPAD        "016"
-    #define __builtin_bswap_wordsize(x) __builtin_bswap64(x)
-#elif (WORD_SIZE==32)
-    typedef uint32_t    word_t;
-    #define ONE         1UL
-    #define MUL_SHIFT   5
-    #define WFMT        "x"
-    #define WPAD        "08"
-    #define __builtin_bswap_wordsize(x) __builtin_bswap32(x)
-#elif (WORD_SIZE==16)
-    typedef uint16_t    word_t;
-    #define ONE         1
-    #define MUL_SHIFT   4
-    #define WFMT        "hx"
-    #define WPAD        "04"
-    #define __builtin_bswap_wordsize(x) __builtin_bswap16(x)
-#elif (WORD_SIZE==8)
-    typedef uint8_t     word_t;
-    #define ONE         1
-    #define MUL_SHIFT   3
-    #define WFMT        "hhx"
-    #define WPAD        "02"
-    #define __builtin_bswap_wordsize(x) (x)
-#else
-#error "invalid word size"
-#endif
-
-void bs_transpose(word_t * blocks);
-void bs_transpose_rev(word_t * blocks);
-void bs_transpose_dst(word_t * transpose, word_t * blocks);
-
-void bs_sbox(word_t U[8]);
-void bs_sbox_rev(word_t U[8]);
-
-void bs_shiftrows(word_t * B);
-void bs_shiftrows_rev(word_t * B);
-
-void bs_mixcolumns(word_t * B);
-void bs_mixcolumns_rev(word_t * B);
-
-void bs_shiftmix(word_t * B);
-
-void bs_addroundkey(word_t * B, word_t * rk);
-void bs_apply_sbox(word_t * input);
-void bs_apply_sbox_rev(word_t * input);
-
-
-void expand_key(unsigned char *in);
-void bs_expand_key(word_t (* rk)[BLOCK_SIZE], uint8_t * key);
-
-void bs_cipher(word_t state[BLOCK_SIZE], word_t (* rk)[BLOCK_SIZE]);
-void bs_cipher_rev(word_t state[BLOCK_SIZE], word_t (* rk)[BLOCK_SIZE]);
-
-
-void dump_hex(uint8_t * h, int len);
-void dump_word(word_t * h, int len);
-void dump_block(word_t * h, int len);
-
-#define MIN(X,Y) ((X) < (Y) ? (X) : (Y))
-#define MAX(X,Y) ((X) > (Y) ? (X) : (Y))
-
-void aes_ecb_encrypt(uint8_t * outputb, uint8_t * inputb, size_t size, uint8_t * key)
-{
-    word_t input_space[BLOCK_SIZE];
-    word_t rk[11][BLOCK_SIZE];
-
-    memset(outputb,0,size);
-    word_t * state = (word_t *)outputb;
-
-    bs_expand_key(rk, key);
-
-    while (size > 0)
-    {
-        if (size < BS_BLOCK_SIZE)
-        {
-            memset(input_space,0,BS_BLOCK_SIZE);
-            memmove(input_space, inputb, size);
-            bs_cipher(input_space,rk);
-            memmove(outputb, input_space, size);
-            size = 0;
-            state += size;
-        }
-        else
-        {
-            memmove(state,inputb,BS_BLOCK_SIZE);
-            bs_cipher(state,rk);
-            size -= BS_BLOCK_SIZE;
-            state += BS_BLOCK_SIZE;
-        }
-
-    }
-}
-
-void aes_ecb_decrypt(uint8_t * outputb, uint8_t * inputb, size_t size, uint8_t * key)
-{
-    word_t input_space[BLOCK_SIZE];
-    word_t rk[11][BLOCK_SIZE];
-
-    memset(outputb,0,size);
-    word_t * state = (word_t *)outputb;
-    
-    bs_expand_key(rk, key);
-
-    while (size > 0)
-    {
-        if (size < BS_BLOCK_SIZE)
-        {
-            memset(input_space,0,BS_BLOCK_SIZE);
-            memmove(input_space, inputb, size);
-            bs_cipher_rev(input_space,rk);
-            memmove(outputb, input_space, size);
-            size = 0;
-            state += size;
-        }
-        else
-        {
-            memmove(state,inputb,BS_BLOCK_SIZE);
-            bs_cipher_rev(state,rk);
-            size -= BS_BLOCK_SIZE;
-            state += BS_BLOCK_SIZE;
-        }
-
-    }
-}
-
-static void INC_CTR(uint8_t * ctr, uint8_t i)
-{
-    ctr += BLOCK_SIZE/8 - 1;
-    uint8_t n = *(ctr);
-    *ctr += i;
-    while(*ctr < n)
-    {
-        ctr--;
-        n = *ctr;
-        (*ctr)++;
-    }
-}
-
-void aes_ctr_encrypt(uint8_t * outputb, uint8_t * inputb, size_t size, uint8_t * key, uint8_t * iv)
-{
-    word_t rk[11][BLOCK_SIZE];
-    word_t ctr[BLOCK_SIZE];
-    uint8_t iv_copy[BLOCK_SIZE/8];
-    
-    memset(outputb,0,size);
-    memset(ctr,0,sizeof(ctr));
-    memmove(iv_copy,iv,BLOCK_SIZE/8);
-
-    word_t * state = (word_t *)outputb;
-    bs_expand_key(rk, key);
-
-    do
-    {
-        int chunk = MIN(size, BS_BLOCK_SIZE);
-        int blocks = chunk / (BLOCK_SIZE/8);
-        if (chunk % (BLOCK_SIZE/8))
-        {
-            blocks++;
-        }
-
-        int i;
-        for (i = 0; i < blocks; i++)
-        {
-            memmove(ctr + (i * WORDS_PER_BLOCK), iv_copy, BLOCK_SIZE/8);
-            INC_CTR(iv_copy,1);
-        }
-
-        bs_cipher(ctr, rk);
-        size -= chunk;
-
-        uint8_t * ctr_p = (uint8_t *) ctr;
-        while(chunk--)
-        {
-            *outputb++ = *ctr_p++ ^ *inputb++;
-        }
-
-    }
-    while(size);
-
-}
-
-void dump_hex(uint8_t * h, int len)
-{
-    while(len--)
-        printf("%02hhx",*h++);
-    printf("\n");
-}
-
-void dump_word(word_t * h, int len)
-{
-    while(len--)
-        if ((len+1) % 8) printf("%" WPAD WFMT "\n",*h++);
-        else printf("%d:\n%" WPAD WFMT "\n",128-len-1,*h++);
-
-    printf("\n");
-}
-
-void dump_block(word_t * h, int len)
-{
-    while(len-=2 >= 0)
-        printf("%" WPAD WFMT"%" WPAD WFMT  "\n",*h++,*h++);
-    printf("\n");
-}
-
-static const uint8_t sbox[256] =   {
-  //0     1    2      3     4    5     6     7      8    9     A      B    C     D     E     F
-  0x63, 0x7c, 0x77, 0x7b, 0xf2, 0x6b, 0x6f, 0xc5, 0x30, 0x01, 0x67, 0x2b, 0xfe, 0xd7, 0xab, 0x76,
-  0xca, 0x82, 0xc9, 0x7d, 0xfa, 0x59, 0x47, 0xf0, 0xad, 0xd4, 0xa2, 0xaf, 0x9c, 0xa4, 0x72, 0xc0,
-  0xb7, 0xfd, 0x93, 0x26, 0x36, 0x3f, 0xf7, 0xcc, 0x34, 0xa5, 0xe5, 0xf1, 0x71, 0xd8, 0x31, 0x15,
-  0x04, 0xc7, 0x23, 0xc3, 0x18, 0x96, 0x05, 0x9a, 0x07, 0x12, 0x80, 0xe2, 0xeb, 0x27, 0xb2, 0x75,
-  0x09, 0x83, 0x2c, 0x1a, 0x1b, 0x6e, 0x5a, 0xa0, 0x52, 0x3b, 0xd6, 0xb3, 0x29, 0xe3, 0x2f, 0x84,
-  0x53, 0xd1, 0x00, 0xed, 0x20, 0xfc, 0xb1, 0x5b, 0x6a, 0xcb, 0xbe, 0x39, 0x4a, 0x4c, 0x58, 0xcf,
-  0xd0, 0xef, 0xaa, 0xfb, 0x43, 0x4d, 0x33, 0x85, 0x45, 0xf9, 0x02, 0x7f, 0x50, 0x3c, 0x9f, 0xa8,
-  0x51, 0xa3, 0x40, 0x8f, 0x92, 0x9d, 0x38, 0xf5, 0xbc, 0xb6, 0xda, 0x21, 0x10, 0xff, 0xf3, 0xd2,
-  0xcd, 0x0c, 0x13, 0xec, 0x5f, 0x97, 0x44, 0x17, 0xc4, 0xa7, 0x7e, 0x3d, 0x64, 0x5d, 0x19, 0x73,
-  0x60, 0x81, 0x4f, 0xdc, 0x22, 0x2a, 0x90, 0x88, 0x46, 0xee, 0xb8, 0x14, 0xde, 0x5e, 0x0b, 0xdb,
-  0xe0, 0x32, 0x3a, 0x0a, 0x49, 0x06, 0x24, 0x5c, 0xc2, 0xd3, 0xac, 0x62, 0x91, 0x95, 0xe4, 0x79,
-  0xe7, 0xc8, 0x37, 0x6d, 0x8d, 0xd5, 0x4e, 0xa9, 0x6c, 0x56, 0xf4, 0xea, 0x65, 0x7a, 0xae, 0x08,
-  0xba, 0x78, 0x25, 0x2e, 0x1c, 0xa6, 0xb4, 0xc6, 0xe8, 0xdd, 0x74, 0x1f, 0x4b, 0xbd, 0x8b, 0x8a,
-  0x70, 0x3e, 0xb5, 0x66, 0x48, 0x03, 0xf6, 0x0e, 0x61, 0x35, 0x57, 0xb9, 0x86, 0xc1, 0x1d, 0x9e,
-  0xe1, 0xf8, 0x98, 0x11, 0x69, 0xd9, 0x8e, 0x94, 0x9b, 0x1e, 0x87, 0xe9, 0xce, 0x55, 0x28, 0xdf,
-  0x8c, 0xa1, 0x89, 0x0d, 0xbf, 0xe6, 0x42, 0x68, 0x41, 0x99, 0x2d, 0x0f, 0xb0, 0x54, 0xbb, 0x16 };
-
-static void rotate(unsigned char *in) {
-    unsigned char a,c;
-    a = in[0];
-    for(c=0;c<3;c++) 
-        in[c] = in[c + 1];
-    in[3] = a;
-    return;
-}
-
-/* Calculate the rcon used in key expansion */
-static unsigned char rcon(unsigned char in) {
-    unsigned char c=1;
-    if(in == 0)  
-        return 0; 
-    while(in != 1) {
-        unsigned char b;
-        b = c & 0x80;
-        c <<= 1;
-        if(b == 0x80) {
-            c ^= 0x1b;
-        }
-        in--;
-    }
-    return c;
-}
-
-/* This is the core key expansion, which, given a 4-byte value,
- * does some scrambling */
-static void schedule_core(unsigned char *in, unsigned char i) {
-    char a;
-    /* Rotate the input 8 bits to the left */
-    rotate(in);
-    /* Apply Rijndael's s-box on all 4 bytes */
-    for(a = 0; a < 4; a++) 
-        in[a] = sbox[in[a]];
-    /* On just the first byte, add 2^i to the byte */
-    in[0] ^= rcon(i);
-}
-
-void expand_key(unsigned char *in) {
-    unsigned char t[4];
-    /* c is 16 because the first sub-key is the user-supplied key */
-    unsigned char c = 16;
-    unsigned char i = 1;
-    unsigned char a;
-
-    /* We need 11 sets of sixteen bytes each for 128-bit mode */
-    while(c < 176) {
-        /* Copy the temporary variable over from the last 4-byte
-         * block */
-        for(a = 0; a < 4; a++) 
-            t[a] = in[a + c - 4];
-        /* Every four blocks (of four bytes), 
-         * do a complex calculation */
-        if(c % 16 == 0) {
-            schedule_core(t,i);
-            i++;
-        }
-        for(a = 0; a < 4; a++) {
-            in[c] = in[c - 16] ^ t[a];
-            c++;
-        }
-    }
-}
-
-#if (defined(__BYTE_ORDER__) && __BYTE_ORDER__ == __ORDER_LITTLE_ENDIAN__) ||\
-        defined(__amd64__) || defined(__amd32__)|| defined(__amd16__)
-#define bs2le(x) (x)
-#define bs2be(x) (x)
-#elif (defined(__BYTE_ORDER__) && __BYTE_ORDER__ == __ORDER_LITTLE_ENDIAN__) ||\
-            (defined(__sparc__))
-#define bs2le(x) __builtin_bswap_wordsize(x)
-#define bs2be(x) __builtin_bswap_wordsize(x)
-#else
-#error "endianness not supported"
-#endif
-
-
-void bs_addroundkey(word_t * B, word_t * rk)
-{
-    int i;
-    for (i = 0; i < BLOCK_SIZE; i++)
-        B[i] ^= rk[i];
-}
-
-void bs_apply_sbox(word_t * input)
-{
-    int i;
-    for(i=0; i < BLOCK_SIZE; i+=8)
-    {
-        bs_sbox(input+i);
-    }
-}
-
-void bs_apply_sbox_rev(word_t * input)
-{
-    int i;
-    for(i=0; i < BLOCK_SIZE; i+=8)
-    {
-        bs_sbox_rev(input+i);
-    }
-}
-
-/*July 2011*/
-/*Straight-line program for AES s box*/
-
-/*Input is U[0], U[1],...,U[7]*/
-/*Output is S[0], S[1],...,S[7]*/
-// http://cs-www.cs.yale.edu/homes/peralta/CircuitStuff/CMT.html
-void bs_sbox_rev(word_t U[8])
-{
-    word_t W[8];
-    word_t
-        T1,T2,T3,T4,T5,T6,T8,
-        T9,T10,T13,T14,T15,T16,
-        T17,T18,T19,T20,T22,T23,T24,
-        T25, T26, T27;
-
-    word_t
-        M1,M2,M3,M4,M5,M6,M7,M8,
-        M9,M10,M11,M12,M13,M14,M15,
-        M16,M17,M18,M19,M20,M21,M22,
-        M23,M24,M25,M26,M27,M28,M29,
-        M30,M31,M32,M33,M34,M35,M36,
-        M37,M38,M39,M40,M41,M42,M43,
-        M44,M45,M46,M47,M48,M49,M50,
-        M51,M52,M53,M54,M55,M56,M57,
-        M58,M59,M60,M61,M62,M63;
-
-    word_t
-        P0,P1,P2,P3,P4,P5,P6,P7,P8,
-        P9,P10,P11,P12,P13,P14,
-        P15,P16,P17,P18,P19,P20,
-        P21,P22,P23,P24,P25,P26,
-        P27,P28,P29;
-
-    word_t Y5,
-        R5, R13, R17, R18, R19;
-
-
-    T23 = U[7] ^ U[4];
-    T22 = ~(U[6] ^ U[4]);
-    T2 = ~(U[7] ^ U[6]);
-    T1 = U[4] ^ U[3];
-    T24 = ~(U[3] ^ U[0]);
-    R5 = U[1] ^ U[0];
-    T8 = ~(U[6] ^ T23);
-    T19 = T22 ^ R5;
-    T9 = ~(U[0] ^ T1);
-    T10 = T2 ^ T24;
-    T13 = T2 ^ R5;
-    T3 = T1 ^ R5;
-    T25 = ~(U[5] ^ T1);
-    R13 = U[6] ^ U[1];
-    T17 = ~(U[5] ^ T19);
-    T20 = T24 ^ R13;
-    T4 = U[3] ^ T8;
-    R17 = ~(U[5] ^ U[2]);
-    R18 = ~(U[2] ^ U[1]);
-    R19 = ~(U[5] ^ U[3]);
-    Y5 = U[7] ^ R17;
-    T6 = T22 ^ R17;
-    T16 = R13 ^ R19;
-    T27 = T1 ^ R18;
-    T15 = T10 ^ T27;
-    T14 = T10 ^ R18;
-    T26 = T3 ^ T16;
-    M1 = T13 & T6;
-    M2 = T23 & T8;
-    M3 = T14 ^ M1;
-    M4 = T19 & Y5;
-    M5 = M4 ^ M1;
-    M6 = T3 & T16;
-    M7 = T22 & T9;
-    M8 = T26 ^ M6;
-    M9 = T20 & T17;
-    M10 = M9 ^ M6;
-    M11 = T1 & T15;
-    M12 = T4 & T27;
-    M13 = M12 ^ M11;
-    M14 = T2 & T10;
-    M15 = M14 ^ M11;
-    M16 = M3 ^ M2;
-    M17 = M5 ^ T24;
-    M18 = M8 ^ M7;
-    M19 = M10 ^ M15;
-    M20 = M16 ^ M13;
-    M21 = M17 ^ M15;
-    M22 = M18 ^ M13;
-    M23 = M19 ^ T25;
-    M24 = M22 ^ M23;
-    M25 = M22 & M20;
-    M26 = M21 ^ M25;
-    M27 = M20 ^ M21;
-    M28 = M23 ^ M25;
-    M29 = M28 & M27;
-    M30 = M26 & M24;
-    M31 = M20 & M23;
-    M32 = M27 & M31;
-    M33 = M27 ^ M25;
-    M34 = M21 & M22;
-    M35 = M24 & M34;
-    M36 = M24 ^ M25;
-    M37 = M21 ^ M29;
-    M38 = M32 ^ M33;
-    M39 = M23 ^ M30;
-    M40 = M35 ^ M36;
-    M41 = M38 ^ M40;
-    M42 = M37 ^ M39;
-    M43 = M37 ^ M38;
-    M44 = M39 ^ M40;
-    M45 = M42 ^ M41;
-    M46 = M44 & T6;
-    M47 = M40 & T8;
-    M48 = M39 & Y5;
-    M49 = M43 & T16;
-    M50 = M38 & T9;
-    M51 = M37 & T17;
-    M52 = M42 & T15;
-    M53 = M45 & T27;
-    M54 = M41 & T10;
-    M55 = M44 & T13;
-    M56 = M40 & T23;
-    M57 = M39 & T19;
-    M58 = M43 & T3;
-    M59 = M38 & T22;
-    M60 = M37 & T20;
-    M61 = M42 & T1;
-    M62 = M45 & T4;
-    M63 = M41 & T2;
-    P0 = M52 ^ M61;
-    P1 = M58 ^ M59;
-    P2 = M54 ^ M62;
-    P3 = M47 ^ M50;
-    P4 = M48 ^ M56;
-    P5 = M46 ^ M51;
-    P6 = M49 ^ M60;
-    P7 = P0 ^ P1;
-    P8 = M50 ^ M53;
-    P9 = M55 ^ M63;
-    P10 = M57 ^ P4;
-    P11 = P0 ^ P3;
-    P12 = M46 ^ M48;
-    P13 = M49 ^ M51;
-    P14 = M49 ^ M62;
-    P15 = M54 ^ M59;
-    P16 = M57 ^ M61;
-    P17 = M58 ^ P2;
-    P18 = M63 ^ P5;
-    P19 = P2 ^ P3;
-    P20 = P4 ^ P6;
-    P22 = P2 ^ P7;
-    P23 = P7 ^ P8;
-    P24 = P5 ^ P7;
-    P25 = P6 ^ P10;
-    P26 = P9 ^ P11;
-    P27 = P10 ^ P18;
-    P28 = P11 ^ P25;
-    P29 = P15 ^ P20;
-    W[7] = P13 ^ P22;
-    W[6] = P26 ^ P29;
-    W[5] = P17 ^ P28;
-    W[4] = P12 ^ P22;
-    W[3] = P23 ^ P27;
-    W[2] = P19 ^ P24;
-    W[1] = P14 ^ P23;
-    W[0] = P9 ^ P16;
-
-    memmove(U,W,sizeof(W));
-}
-
-void bs_sbox(word_t U[8])
-{
-    word_t S[8];
-    word_t
-        T1,T2,T3,T4,T5,T6,T7,T8,
-        T9,T10,T11,T12,T13,T14,T15,T16,
-        T17,T18,T19,T20,T21,T22,T23,T24,
-        T25, T26, T27;
-
-    word_t
-        M1,M2,M3,M4,M5,M6,M7,M8,
-        M9,M10,M11,M12,M13,M14,M15,
-        M16,M17,M18,M19,M20,M21,M22,
-        M23,M24,M25,M26,M27,M28,M29,
-        M30,M31,M32,M33,M34,M35,M36,
-        M37,M38,M39,M40,M41,M42,M43,
-        M44,M45,M46,M47,M48,M49,M50,
-        M51,M52,M53,M54,M55,M56,M57,
-        M58,M59,M60,M61,M62,M63;
-
-    word_t
-        L0,L1,L2,L3,L4,L5,L6,L7,L8,
-        L9,L10,L11,L12,L13,L14,
-        L15,L16,L17,L18,L19,L20,
-        L21,L22,L23,L24,L25,L26,
-        L27,L28,L29;
-
-    T1 = U[7] ^ U[4];
-    T2 = U[7] ^ U[2];
-    T3 = U[7] ^ U[1];
-    T4 = U[4] ^ U[2];
-    T5 = U[3] ^ U[1];
-    T6 = T1 ^ T5;
-    T7 = U[6] ^ U[5];
-    T8 = U[0] ^ T6;
-    T9 = U[0] ^ T7;
-    T10 = T6 ^ T7;
-    T11 = U[6] ^ U[2];
-    T12 = U[5] ^ U[2];
-    T13 = T3 ^ T4;
-    T14 = T6 ^ T11;
-    T15 = T5 ^ T11;
-    T16 = T5 ^ T12;
-    T17 = T9 ^ T16;
-    T18 = U[4] ^ U[0];
-    T19 = T7 ^ T18;
-    T20 = T1 ^ T19;
-    T21 = U[1] ^ U[0];
-    T22 = T7 ^ T21;
-    T23 = T2 ^ T22;
-    T24 = T2 ^ T10;
-    T25 = T20 ^ T17;
-    T26 = T3 ^ T16;
-    T27 = T1 ^ T12;
-    M1 = T13 & T6;
-    M2 = T23 & T8;
-    M3 = T14 ^ M1;
-    M4 = T19 & U[0];
-    M5 = M4 ^ M1;
-    M6 = T3 & T16;
-    M7 = T22 & T9;
-    M8 = T26 ^ M6;
-    M9 = T20 & T17;
-    M10 = M9 ^ M6;
-    M11 = T1 & T15;
-    M12 = T4 & T27;
-    M13 = M12 ^ M11;
-    M14 = T2 & T10;
-    M15 = M14 ^ M11;
-    M16 = M3 ^ M2;
-    M17 = M5 ^ T24;
-    M18 = M8 ^ M7;
-    M19 = M10 ^ M15;
-    M20 = M16 ^ M13;
-    M21 = M17 ^ M15;
-    M22 = M18 ^ M13;
-    M23 = M19 ^ T25;
-    M24 = M22 ^ M23;
-    M25 = M22 & M20;
-    M26 = M21 ^ M25;
-    M27 = M20 ^ M21;
-    M28 = M23 ^ M25;
-    M29 = M28 & M27;
-    M30 = M26 & M24;
-    M31 = M20 & M23;
-    M32 = M27 & M31;
-    M33 = M27 ^ M25;
-    M34 = M21 & M22;
-    M35 = M24 & M34;
-    M36 = M24 ^ M25;
-    M37 = M21 ^ M29;
-    M38 = M32 ^ M33;
-    M39 = M23 ^ M30;
-    M40 = M35 ^ M36;
-    M41 = M38 ^ M40;
-    M42 = M37 ^ M39;
-    M43 = M37 ^ M38;
-    M44 = M39 ^ M40;
-    M45 = M42 ^ M41;
-    M46 = M44 & T6;
-    M47 = M40 & T8;
-    M48 = M39 & U[0];
-    M49 = M43 & T16;
-    M50 = M38 & T9;
-    M51 = M37 & T17;
-    M52 = M42 & T15;
-    M53 = M45 & T27;
-    M54 = M41 & T10;
-    M55 = M44 & T13;
-    M56 = M40 & T23;
-    M57 = M39 & T19;
-    M58 = M43 & T3;
-    M59 = M38 & T22;
-    M60 = M37 & T20;
-    M61 = M42 & T1;
-    M62 = M45 & T4;
-    M63 = M41 & T2;
-    L0 = M61 ^ M62;
-    L1 = M50 ^ M56;
-    L2 = M46 ^ M48;
-    L3 = M47 ^ M55;
-    L4 = M54 ^ M58;
-    L5 = M49 ^ M61;
-    L6 = M62 ^ L5;
-    L7 = M46 ^ L3;
-    L8 = M51 ^ M59;
-    L9 = M52 ^ M53;
-    L10 = M53 ^ L4;
-    L11 = M60 ^ L2;
-    L12 = M48 ^ M51;
-    L13 = M50 ^ L0;
-    L14 = M52 ^ M61;
-    L15 = M55 ^ L1;
-    L16 = M56 ^ L0;
-    L17 = M57 ^ L1;
-    L18 = M58 ^ L8;
-    L19 = M63 ^ L4;
-    L20 = L0 ^ L1;
-    L21 = L1 ^ L7;
-    L22 = L3 ^ L12;
-    L23 = L18 ^ L2;
-    L24 = L15 ^ L9;
-    L25 = L6 ^ L10;
-    L26 = L7 ^ L9;
-    L27 = L8 ^ L10;
-    L28 = L11 ^ L14;
-    L29 = L11 ^ L17;
-    S[7] = L6 ^ L24;
-    S[6] = ~(L16 ^ L26);
-    S[5] = ~(L19 ^ L28);
-    S[4] = L6 ^ L21;
-    S[3] = L20 ^ L22;
-    S[2] = L25 ^ L29;
-    S[1] = ~(L13 ^ L27);
-    S[0] = ~(L6 ^ L23);
-
-    memmove(U,S,sizeof(S));
-}
-
-void bs_transpose(word_t * blocks)
-{
-    word_t transpose[BLOCK_SIZE];
-    memset(transpose, 0, sizeof(transpose));
-    bs_transpose_dst(transpose,blocks);
-    memmove(blocks,transpose,sizeof(transpose));
-}
-
-void bs_transpose_dst(word_t * transpose, word_t * blocks)
-{
-    int i,k;
-    word_t w;
-    for(k=0; k < WORD_SIZE; k++)
-    {
-        int bitpos = ONE << k;
-        for (i=0; i < WORDS_PER_BLOCK; i++)
-        {
-            w = bs2le(blocks[k * WORDS_PER_BLOCK + i]);
-            int offset = i << MUL_SHIFT;
-
-#ifndef UNROLL_TRANSPOSE
-            int j;
-            for(j=0; j < WORD_SIZE; j++)
-            {
-                // TODO make const time
-                transpose[offset + j] |= (w & (ONE << j)) ? bitpos : 0;
-            }
-#else
-
-            transpose[(offset)+ 0 ] |= (w & (ONE << 0 )) ? (bitpos) : 0;
-            transpose[(offset)+ 1 ] |= (w & (ONE << 1 )) ? (bitpos) : 0;
-            transpose[(offset)+ 2 ] |= (w & (ONE << 2 )) ? (bitpos) : 0;
-            transpose[(offset)+ 3 ] |= (w & (ONE << 3 )) ? (bitpos) : 0;
-            transpose[(offset)+ 4 ] |= (w & (ONE << 4 )) ? (bitpos) : 0;
-            transpose[(offset)+ 5 ] |= (w & (ONE << 5 )) ? (bitpos) : 0;
-            transpose[(offset)+ 6 ] |= (w & (ONE << 6 )) ? (bitpos) : 0;
-            transpose[(offset)+ 7 ] |= (w & (ONE << 7 )) ? (bitpos) : 0;
-#if WORD_SIZE > 8
-            transpose[(offset)+ 8 ] |= (w & (ONE << 8 )) ? (bitpos) : 0;
-            transpose[(offset)+ 9 ] |= (w & (ONE << 9 )) ? (bitpos) : 0;
-            transpose[(offset)+ 10] |= (w & (ONE << 10)) ? (bitpos) : 0;
-            transpose[(offset)+ 11] |= (w & (ONE << 11)) ? (bitpos) : 0;
-            transpose[(offset)+ 12] |= (w & (ONE << 12)) ? (bitpos) : 0;
-            transpose[(offset)+ 13] |= (w & (ONE << 13)) ? (bitpos) : 0;
-            transpose[(offset)+ 14] |= (w & (ONE << 14)) ? (bitpos) : 0;
-            transpose[(offset)+ 15] |= (w & (ONE << 15)) ? (bitpos) : 0;
-#endif
-#if WORD_SIZE > 16
-            transpose[(offset)+ 16] |= (w & (ONE << 16)) ? (bitpos) : 0;
-            transpose[(offset)+ 17] |= (w & (ONE << 17)) ? (bitpos) : 0;
-            transpose[(offset)+ 18] |= (w & (ONE << 18)) ? (bitpos) : 0;
-            transpose[(offset)+ 19] |= (w & (ONE << 19)) ? (bitpos) : 0;
-            transpose[(offset)+ 20] |= (w & (ONE << 20)) ? (bitpos) : 0;
-            transpose[(offset)+ 21] |= (w & (ONE << 21)) ? (bitpos) : 0;
-            transpose[(offset)+ 22] |= (w & (ONE << 22)) ? (bitpos) : 0;
-            transpose[(offset)+ 23] |= (w & (ONE << 23)) ? (bitpos) : 0;
-            transpose[(offset)+ 24] |= (w & (ONE << 24)) ? (bitpos) : 0;
-            transpose[(offset)+ 25] |= (w & (ONE << 25)) ? (bitpos) : 0;
-            transpose[(offset)+ 26] |= (w & (ONE << 26)) ? (bitpos) : 0;
-            transpose[(offset)+ 27] |= (w & (ONE << 27)) ? (bitpos) : 0;
-            transpose[(offset)+ 28] |= (w & (ONE << 28)) ? (bitpos) : 0;
-            transpose[(offset)+ 29] |= (w & (ONE << 29)) ? (bitpos) : 0;
-            transpose[(offset)+ 30] |= (w & (ONE << 30)) ? (bitpos) : 0;
-            transpose[(offset)+ 31] |= (w & (ONE << 31)) ? (bitpos) : 0;
-#endif
-#if WORD_SIZE > 32
-            transpose[(offset)+ 32] |= (w & (ONE << 32)) ? (bitpos) : 0;
-            transpose[(offset)+ 33] |= (w & (ONE << 33)) ? (bitpos) : 0;
-            transpose[(offset)+ 34] |= (w & (ONE << 34)) ? (bitpos) : 0;
-            transpose[(offset)+ 35] |= (w & (ONE << 35)) ? (bitpos) : 0;
-            transpose[(offset)+ 36] |= (w & (ONE << 36)) ? (bitpos) : 0;
-            transpose[(offset)+ 37] |= (w & (ONE << 37)) ? (bitpos) : 0;
-            transpose[(offset)+ 38] |= (w & (ONE << 38)) ? (bitpos) : 0;
-            transpose[(offset)+ 39] |= (w & (ONE << 39)) ? (bitpos) : 0;
-            transpose[(offset)+ 40] |= (w & (ONE << 40)) ? (bitpos) : 0;
-            transpose[(offset)+ 41] |= (w & (ONE << 41)) ? (bitpos) : 0;
-            transpose[(offset)+ 42] |= (w & (ONE << 42)) ? (bitpos) : 0;
-            transpose[(offset)+ 43] |= (w & (ONE << 43)) ? (bitpos) : 0;
-            transpose[(offset)+ 44] |= (w & (ONE << 44)) ? (bitpos) : 0;
-            transpose[(offset)+ 45] |= (w & (ONE << 45)) ? (bitpos) : 0;
-            transpose[(offset)+ 46] |= (w & (ONE << 46)) ? (bitpos) : 0;
-            transpose[(offset)+ 47] |= (w & (ONE << 47)) ? (bitpos) : 0;
-            transpose[(offset)+ 48] |= (w & (ONE << 48)) ? (bitpos) : 0;
-            transpose[(offset)+ 49] |= (w & (ONE << 49)) ? (bitpos) : 0;
-            transpose[(offset)+ 50] |= (w & (ONE << 50)) ? (bitpos) : 0;
-            transpose[(offset)+ 51] |= (w & (ONE << 51)) ? (bitpos) : 0;
-            transpose[(offset)+ 52] |= (w & (ONE << 52)) ? (bitpos) : 0;
-            transpose[(offset)+ 53] |= (w & (ONE << 53)) ? (bitpos) : 0;
-            transpose[(offset)+ 54] |= (w & (ONE << 54)) ? (bitpos) : 0;
-            transpose[(offset)+ 55] |= (w & (ONE << 55)) ? (bitpos) : 0;
-            transpose[(offset)+ 56] |= (w & (ONE << 56)) ? (bitpos) : 0;
-            transpose[(offset)+ 57] |= (w & (ONE << 57)) ? (bitpos) : 0;
-            transpose[(offset)+ 58] |= (w & (ONE << 58)) ? (bitpos) : 0;
-            transpose[(offset)+ 59] |= (w & (ONE << 59)) ? (bitpos) : 0;
-            transpose[(offset)+ 60] |= (w & (ONE << 60)) ? (bitpos) : 0;
-            transpose[(offset)+ 61] |= (w & (ONE << 61)) ? (bitpos) : 0;
-            transpose[(offset)+ 62] |= (w & (ONE << 62)) ? (bitpos) : 0;
-            transpose[(offset)+ 63] |= (w & (ONE << 63)) ? (bitpos) : 0;
-#endif
-#endif
-                // constant time:
-                //transpose[(i<<MUL_SHIFT)+ j] |= (((int64_t)((w & (ONE << j)) << (WORD_SIZE-1-j)))>>(WORD_SIZE-1)) & (ONE<<k);
-        }
-    }
-}
-
-void bs_transpose_rev(word_t * blocks)
-{
-    int i,k;
-    word_t w;
-    word_t transpose[BLOCK_SIZE];
-    memset(transpose, 0, sizeof(transpose));
-    for(k=0; k < BLOCK_SIZE; k++)
-    {
-        w = blocks[k];
-        word_t bitpos = bs2be(ONE << (k % WORD_SIZE));
-        word_t offset = k / WORD_SIZE;
-#ifndef UNROLL_TRANSPOSE
-        int j;
-        for(j=0; j < WORD_SIZE; j++)
-        {
-            word_t bit = (w & (ONE << j)) ? (ONE << (k % WORD_SIZE)) : 0;
-            transpose[j * WORDS_PER_BLOCK + (offset)] |= bit;
-        }
-#else
-        transpose[0  * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 0 )) ? bitpos : 0;
-        transpose[1  * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 1 )) ? bitpos : 0;
-        transpose[2  * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 2 )) ? bitpos : 0;
-        transpose[3  * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 3 )) ? bitpos : 0;
-        transpose[4  * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 4 )) ? bitpos : 0;
-        transpose[5  * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 5 )) ? bitpos : 0;
-        transpose[6  * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 6 )) ? bitpos : 0;
-        transpose[7  * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 7 )) ? bitpos : 0;
-#if WORD_SIZE > 8
-        transpose[8  * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 8 )) ? bitpos : 0;
-        transpose[9  * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 9 )) ? bitpos : 0;
-        transpose[10 * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 10)) ? bitpos : 0;
-        transpose[11 * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 11)) ? bitpos : 0;
-        transpose[12 * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 12)) ? bitpos : 0;
-        transpose[13 * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 13)) ? bitpos : 0;
-        transpose[14 * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 14)) ? bitpos : 0;
-        transpose[15 * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 15)) ? bitpos : 0;
-#endif
-#if WORD_SIZE > 16
-        transpose[16 * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 16)) ? bitpos : 0;
-        transpose[17 * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 17)) ? bitpos : 0;
-        transpose[18 * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 18)) ? bitpos : 0;
-        transpose[19 * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 19)) ? bitpos : 0;
-        transpose[20 * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 20)) ? bitpos : 0;
-        transpose[21 * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 21)) ? bitpos : 0;
-        transpose[22 * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 22)) ? bitpos : 0;
-        transpose[23 * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 23)) ? bitpos : 0;
-        transpose[24 * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 24)) ? bitpos : 0;
-        transpose[25 * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 25)) ? bitpos : 0;
-        transpose[26 * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 26)) ? bitpos : 0;
-        transpose[27 * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 27)) ? bitpos : 0;
-        transpose[28 * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 28)) ? bitpos : 0;
-        transpose[29 * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 29)) ? bitpos : 0;
-        transpose[30 * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 30)) ? bitpos : 0;
-        transpose[31 * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 31)) ? bitpos : 0;
-#endif
-#if WORD_SIZE > 32
-        transpose[32 * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 32)) ? bitpos : 0;
-        transpose[33 * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 33)) ? bitpos : 0;
-        transpose[34 * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 34)) ? bitpos : 0;
-        transpose[35 * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 35)) ? bitpos : 0;
-        transpose[36 * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 36)) ? bitpos : 0;
-        transpose[37 * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 37)) ? bitpos : 0;
-        transpose[38 * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 38)) ? bitpos : 0;
-        transpose[39 * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 39)) ? bitpos : 0;
-        transpose[40 * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 40)) ? bitpos : 0;
-        transpose[41 * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 41)) ? bitpos : 0;
-        transpose[42 * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 42)) ? bitpos : 0;
-        transpose[43 * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 43)) ? bitpos : 0;
-        transpose[44 * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 44)) ? bitpos : 0;
-        transpose[45 * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 45)) ? bitpos : 0;
-        transpose[46 * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 46)) ? bitpos : 0;
-        transpose[47 * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 47)) ? bitpos : 0;
-        transpose[48 * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 48)) ? bitpos : 0;
-        transpose[49 * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 49)) ? bitpos : 0;
-        transpose[50 * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 50)) ? bitpos : 0;
-        transpose[51 * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 51)) ? bitpos : 0;
-        transpose[52 * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 52)) ? bitpos : 0;
-        transpose[53 * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 53)) ? bitpos : 0;
-        transpose[54 * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 54)) ? bitpos : 0;
-        transpose[55 * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 55)) ? bitpos : 0;
-        transpose[56 * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 56)) ? bitpos : 0;
-        transpose[57 * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 57)) ? bitpos : 0;
-        transpose[58 * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 58)) ? bitpos : 0;
-        transpose[59 * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 59)) ? bitpos : 0;
-        transpose[60 * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 60)) ? bitpos : 0;
-        transpose[61 * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 61)) ? bitpos : 0;
-        transpose[62 * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 62)) ? bitpos : 0;
-        transpose[63 * WORDS_PER_BLOCK + (offset )] |= (w & (ONE << 63)) ? bitpos : 0;
-#endif
-#endif
-    }
-    memmove(blocks,transpose,sizeof(transpose));
-}
-
-
-#define R0          0
-#define R1          8
-#define R2          16
-#define R3          24
-
-#define B0          0
-#define B1          32
-#define B2          64
-#define B3          96
-
-#define R0_shift        (BLOCK_SIZE/4)*0
-#define R1_shift        (BLOCK_SIZE/4)*1
-#define R2_shift        (BLOCK_SIZE/4)*2
-#define R3_shift        (BLOCK_SIZE/4)*3
-#define B_MOD           (BLOCK_SIZE)
-
-
-void bs_shiftrows(word_t * B)
-{
-    word_t Bp_space[BLOCK_SIZE];
-    word_t * Bp = Bp_space;
-    word_t * Br0 = B + 0;
-    word_t * Br1 = B + 32;
-    word_t * Br2 = B + 64;
-    word_t * Br3 = B + 96;
-    uint8_t offsetr0 = 0;
-    uint8_t offsetr1 = 32;
-    uint8_t offsetr2 = 64;
-    uint8_t offsetr3 = 96;
-
-
-    int i;
-    for(i=0; i<4; i++)
-    {
-        Bp[B0 + 0] = Br0[0];
-        Bp[B0 + 1] = Br0[1];
-        Bp[B0 + 2] = Br0[2];
-        Bp[B0 + 3] = Br0[3];
-        Bp[B0 + 4] = Br0[4];
-        Bp[B0 + 5] = Br0[5];
-        Bp[B0 + 6] = Br0[6];
-        Bp[B0 + 7] = Br0[7];
-        Bp[B1 + 0] = Br1[0];
-        Bp[B1 + 1] = Br1[1];
-        Bp[B1 + 2] = Br1[2];
-        Bp[B1 + 3] = Br1[3];
-        Bp[B1 + 4] = Br1[4];
-        Bp[B1 + 5] = Br1[5];
-        Bp[B1 + 6] = Br1[6];
-        Bp[B1 + 7] = Br1[7];
-        Bp[B2 + 0] = Br2[0];
-        Bp[B2 + 1] = Br2[1];
-        Bp[B2 + 2] = Br2[2];
-        Bp[B2 + 3] = Br2[3];
-        Bp[B2 + 4] = Br2[4];
-        Bp[B2 + 5] = Br2[5];
-        Bp[B2 + 6] = Br2[6];
-        Bp[B2 + 7] = Br2[7];
-        Bp[B3 + 0] = Br3[0];
-        Bp[B3 + 1] = Br3[1];
-        Bp[B3 + 2] = Br3[2];
-        Bp[B3 + 3] = Br3[3];
-        Bp[B3 + 4] = Br3[4];
-        Bp[B3 + 5] = Br3[5];
-        Bp[B3 + 6] = Br3[6];
-        Bp[B3 + 7] = Br3[7];
-
-        offsetr0 = (offsetr0 + BLOCK_SIZE/16 + BLOCK_SIZE/4) & 0x7f;
-        offsetr1 = (offsetr1 + BLOCK_SIZE/16 + BLOCK_SIZE/4) & 0x7f;
-        offsetr2 = (offsetr2 + BLOCK_SIZE/16 + BLOCK_SIZE/4) & 0x7f;
-        offsetr3 = (offsetr3 + BLOCK_SIZE/16 + BLOCK_SIZE/4) & 0x7f;
-
-        Br0 = B + offsetr0;
-        Br1 = B + offsetr1;
-        Br2 = B + offsetr2;
-        Br3 = B + offsetr3;
-
-        Bp += 8;
-    }
-    memmove(B,Bp_space,sizeof(Bp_space));
-}
-
-
-void bs_shiftrows_rev(word_t * B)
-{
-    word_t Bp_space[BLOCK_SIZE];
-    word_t * Bp = Bp_space;
-    word_t * Br0 = Bp + 0;
-    word_t * Br1 = Bp + 32;
-    word_t * Br2 = Bp + 64;
-    word_t * Br3 = Bp + 96;
-    uint8_t offsetr0 = 0;
-    uint8_t offsetr1 = 32;
-    uint8_t offsetr2 = 64;
-    uint8_t offsetr3 = 96;
-
-
-    int i;
-    for(i=0; i<4; i++)
-    {
-        Br0[0] = B[B0 + 0];
-        Br0[1] = B[B0 + 1];
-        Br0[2] = B[B0 + 2];
-        Br0[3] = B[B0 + 3];
-        Br0[4] = B[B0 + 4];
-        Br0[5] = B[B0 + 5];
-        Br0[6] = B[B0 + 6];
-        Br0[7] = B[B0 + 7];
-        Br1[0] = B[B1 + 0];
-        Br1[1] = B[B1 + 1];
-        Br1[2] = B[B1 + 2];
-        Br1[3] = B[B1 + 3];
-        Br1[4] = B[B1 + 4];
-        Br1[5] = B[B1 + 5];
-        Br1[6] = B[B1 + 6];
-        Br1[7] = B[B1 + 7];
-        Br2[0] = B[B2 + 0];
-        Br2[1] = B[B2 + 1];
-        Br2[2] = B[B2 + 2];
-        Br2[3] = B[B2 + 3];
-        Br2[4] = B[B2 + 4];
-        Br2[5] = B[B2 + 5];
-        Br2[6] = B[B2 + 6];
-        Br2[7] = B[B2 + 7];
-        Br3[0] = B[B3 + 0];
-        Br3[1] = B[B3 + 1];
-        Br3[2] = B[B3 + 2];
-        Br3[3] = B[B3 + 3];
-        Br3[4] = B[B3 + 4];
-        Br3[5] = B[B3 + 5];
-        Br3[6] = B[B3 + 6];
-        Br3[7] = B[B3 + 7];
-
-        offsetr0 = (offsetr0 + BLOCK_SIZE/16 + BLOCK_SIZE/4) & 0x7f;
-        offsetr1 = (offsetr1 + BLOCK_SIZE/16 + BLOCK_SIZE/4) & 0x7f;
-        offsetr2 = (offsetr2 + BLOCK_SIZE/16 + BLOCK_SIZE/4) & 0x7f;
-        offsetr3 = (offsetr3 + BLOCK_SIZE/16 + BLOCK_SIZE/4) & 0x7f;
-
-        Br0 = Bp + offsetr0;
-        Br1 = Bp + offsetr1;
-        Br2 = Bp + offsetr2;
-        Br3 = Bp + offsetr3;
-
-        B += 8;
-    }
-    memmove(B - 8 * 4,Bp_space,sizeof(Bp_space));
-}
-
-
-#define A0  0
-#define A1  8
-#define A2  16
-#define A3  24
-
-// Does shift rows and mix columns in same step
-void bs_shiftmix(word_t * B)
-{
-    word_t Bp_space[BLOCK_SIZE];
-    word_t * Bp = Bp_space;
-
-    word_t * Br0 = B + 0;
-    word_t * Br1 = B + 32;
-    word_t * Br2 = B + 64;
-    word_t * Br3 = B + 96;
-
-    uint8_t offsetr0 = 0;
-    uint8_t offsetr1 = 32;
-    uint8_t offsetr2 = 64;
-    uint8_t offsetr3 = 96;
-
-        Br0 = B + offsetr0;
-        Br1 = B + offsetr1;
-        Br2 = B + offsetr2;
-        Br3 = B + offsetr3;
-
-
-    int i;
-    for (i = 0; i < 4; i++)
-    {
-        // B0
-        //            2*A0        2*A1              A1           A2           A3
-        word_t of =Br0[R0+7]^ Br1[R1+7];
-        Bp[A0+0] =                         Br1[R1+0] ^ Br2[R2+0] ^ Br3[R3+0] ^ of;
-        Bp[A0+1] = Br0[R0+0] ^ Br1[R1+0] ^ Br1[R1+1] ^ Br2[R2+1] ^ Br3[R3+1] ^ of;
-        Bp[A0+2] = Br0[R0+1] ^ Br1[R1+1] ^ Br1[R1+2] ^ Br2[R2+2] ^ Br3[R3+2];
-        Bp[A0+3] = Br0[R0+2] ^ Br1[R1+2] ^ Br1[R1+3] ^ Br2[R2+3] ^ Br3[R3+3] ^ of;
-        Bp[A0+4] = Br0[R0+3] ^ Br1[R1+3] ^ Br1[R1+4] ^ Br2[R2+4] ^ Br3[R3+4] ^ of;
-        Bp[A0+5] = Br0[R0+4] ^ Br1[R1+4] ^ Br1[R1+5] ^ Br2[R2+5] ^ Br3[R3+5];
-        Bp[A0+6] = Br0[R0+5] ^ Br1[R1+5] ^ Br1[R1+6] ^ Br2[R2+6] ^ Br3[R3+6];
-        Bp[A0+7] = Br0[R0+6] ^ Br1[R1+6] ^ Br1[R1+7] ^ Br2[R2+7] ^ Br3[R3+7];
-
-        //            A0            2*A1        2*A2        A2       A3
-        of = Br1[R1+7] ^ Br2[R2+7];
-        Bp[A1+0] = Br0[R0+0]                         ^ Br2[R2+0] ^ Br3[R3+0] ^ of;
-        Bp[A1+1] = Br0[R0+1] ^ Br1[R1+0] ^ Br2[R2+0] ^ Br2[R2+1] ^ Br3[R3+1] ^ of;
-        Bp[A1+2] = Br0[R0+2] ^ Br1[R1+1] ^ Br2[R2+1] ^ Br2[R2+2] ^ Br3[R3+2];
-        Bp[A1+3] = Br0[R0+3] ^ Br1[R1+2] ^ Br2[R2+2] ^ Br2[R2+3] ^ Br3[R3+3] ^ of;
-        Bp[A1+4] = Br0[R0+4] ^ Br1[R1+3] ^ Br2[R2+3] ^ Br2[R2+4] ^ Br3[R3+4] ^ of;
-        Bp[A1+5] = Br0[R0+5] ^ Br1[R1+4] ^ Br2[R2+4] ^ Br2[R2+5] ^ Br3[R3+5];
-        Bp[A1+6] = Br0[R0+6] ^ Br1[R1+5] ^ Br2[R2+5] ^ Br2[R2+6] ^ Br3[R3+6];
-        Bp[A1+7] = Br0[R0+7] ^ Br1[R1+6] ^ Br2[R2+6] ^ Br2[R2+7] ^ Br3[R3+7];
-
-        //            A0             A1      2*A2        2*A3         A3
-        of = Br2[R2+7] ^ Br3[R3+7];
-        Bp[A2+0] = Br0[R0+0] ^ Br1[R1+0]                         ^ Br3[R3+0] ^ of;
-        Bp[A2+1] = Br0[R0+1] ^ Br1[R1+1] ^ Br2[R2+0] ^ Br3[R3+0] ^ Br3[R3+1] ^ of;
-        Bp[A2+2] = Br0[R0+2] ^ Br1[R1+2] ^ Br2[R2+1] ^ Br3[R3+1] ^ Br3[R3+2];
-        Bp[A2+3] = Br0[R0+3] ^ Br1[R1+3] ^ Br2[R2+2] ^ Br3[R3+2] ^ Br3[R3+3] ^ of;
-        Bp[A2+4] = Br0[R0+4] ^ Br1[R1+4] ^ Br2[R2+3] ^ Br3[R3+3] ^ Br3[R3+4] ^ of;
-        Bp[A2+5] = Br0[R0+5] ^ Br1[R1+5] ^ Br2[R2+4] ^ Br3[R3+4] ^ Br3[R3+5];
-        Bp[A2+6] = Br0[R0+6] ^ Br1[R1+6] ^ Br2[R2+5] ^ Br3[R3+5] ^ Br3[R3+6];
-        Bp[A2+7] = Br0[R0+7] ^ Br1[R1+7] ^ Br2[R2+6] ^ Br3[R3+6] ^ Br3[R3+7];
-
-        //             A0          2*A0           A1       A2      2*A3
-        of = Br0[R0+7] ^ Br3[R3+7];
-        Bp[A3+0] = Br0[R0+0] ^             Br1[R1+0] ^ Br2[R2+0]             ^ of;
-        Bp[A3+1] = Br0[R0+1] ^ Br0[R0+0] ^ Br1[R1+1] ^ Br2[R2+1] ^ Br3[R3+0] ^ of;
-        Bp[A3+2] = Br0[R0+2] ^ Br0[R0+1] ^ Br1[R1+2] ^ Br2[R2+2] ^ Br3[R3+1];
-        Bp[A3+3] = Br0[R0+3] ^ Br0[R0+2] ^ Br1[R1+3] ^ Br2[R2+3] ^ Br3[R3+2] ^ of;
-        Bp[A3+4] = Br0[R0+4] ^ Br0[R0+3] ^ Br1[R1+4] ^ Br2[R2+4] ^ Br3[R3+3] ^ of;
-        Bp[A3+5] = Br0[R0+5] ^ Br0[R0+4] ^ Br1[R1+5] ^ Br2[R2+5] ^ Br3[R3+4];
-        Bp[A3+6] = Br0[R0+6] ^ Br0[R0+5] ^ Br1[R1+6] ^ Br2[R2+6] ^ Br3[R3+5];
-        Bp[A3+7] = Br0[R0+7] ^ Br0[R0+6] ^ Br1[R1+7] ^ Br2[R2+7] ^ Br3[R3+6];
-
-        Bp += BLOCK_SIZE/4;
-
-        offsetr0 = (offsetr0 + BLOCK_SIZE/4) & 0x7f;
-        offsetr1 = (offsetr1 + BLOCK_SIZE/4) & 0x7f;
-        offsetr2 = (offsetr2 + BLOCK_SIZE/4) & 0x7f;
-        offsetr3 = (offsetr3 + BLOCK_SIZE/4) & 0x7f;
-
-        Br0 = B + offsetr0;
-        Br1 = B + offsetr1;
-        Br2 = B + offsetr2;
-        Br3 = B + offsetr3;
-    }
-
-    memmove(B,Bp_space,sizeof(Bp_space));
-}
-
-
-
-void bs_mixcolumns(word_t * B)
-{
-    word_t Bp_space[BLOCK_SIZE];
-    word_t * Bp = Bp_space;
-    // to understand this, see
-    // https://en.wikipedia.org/wiki/Rijndael_mix_columns
-    
-    int i = 0;
-    for (; i < 4; i++)
-    {
-    //  of = A0 ^ A1;
-    //  A0 = A0 ^ (0x1b & ((signed char)of>>7));
-
-    //// 2 * A0
-    //  A0 = A0 ^ (A0 << 1)
-
-    //// + 3 * A1
-    //  A0 = A0 ^ (A1)
-    //  A0 = A0 ^ (A1<<1)
-
-    //// + A2 + A3
-    //  A0 = A0 ^ (A2)
-    //  A0 = A0 ^ (A3)
-    //          A0.7    A1.7
-    word_t of = B[A0+7] ^ B[A1+7];
-
-    //          2*A0     2*A1          A1      A2           A3
-    Bp[A0+0] =                     B[A1+0] ^ B[A2+0] ^ B[A3+0] ^ of;
-    Bp[A0+1] = B[A0+0] ^ B[A1+0] ^ B[A1+1] ^ B[A2+1] ^ B[A3+1] ^ of;
-    Bp[A0+2] = B[A0+1] ^ B[A1+1] ^ B[A1+2] ^ B[A2+2] ^ B[A3+2];
-    Bp[A0+3] = B[A0+2] ^ B[A1+2] ^ B[A1+3] ^ B[A2+3] ^ B[A3+3] ^ of;
-    Bp[A0+4] = B[A0+3] ^ B[A1+3] ^ B[A1+4] ^ B[A2+4] ^ B[A3+4] ^ of;
-    Bp[A0+5] = B[A0+4] ^ B[A1+4] ^ B[A1+5] ^ B[A2+5] ^ B[A3+5];
-    Bp[A0+6] = B[A0+5] ^ B[A1+5] ^ B[A1+6] ^ B[A2+6] ^ B[A3+6];
-    Bp[A0+7] = B[A0+6] ^ B[A1+6] ^ B[A1+7] ^ B[A2+7] ^ B[A3+7];
-
-
-
-    //  of = A1 ^ A2
-    //  A1 = A1 ^ (0x1b & ((signed char)of>>7));
-
-    //// A0
-    //  A1 = A1 ^ (A0)
-
-    //// + 2 * A1
-    //  A1 = A1 ^ (A1 << 1)
-
-    //// + 3 * A2
-    //  A1 = A1 ^ (A2)
-    //  A1 = A1 ^ (A2<<1)
-
-    //// + A3
-    //  A1 = A1 ^ (A3)
-
-    of = B[A1+7] ^ B[A2+7];
-
-    //          A0      2*A1        2*A2      A2        A3
-    Bp[A1+0] = B[A0+0]                     ^ B[A2+0] ^ B[A3+0] ^ of;
-    Bp[A1+1] = B[A0+1] ^ B[A1+0] ^ B[A2+0] ^ B[A2+1] ^ B[A3+1] ^ of;
-    Bp[A1+2] = B[A0+2] ^ B[A1+1] ^ B[A2+1] ^ B[A2+2] ^ B[A3+2];
-    Bp[A1+3] = B[A0+3] ^ B[A1+2] ^ B[A2+2] ^ B[A2+3] ^ B[A3+3] ^ of;
-    Bp[A1+4] = B[A0+4] ^ B[A1+3] ^ B[A2+3] ^ B[A2+4] ^ B[A3+4] ^ of;
-    Bp[A1+5] = B[A0+5] ^ B[A1+4] ^ B[A2+4] ^ B[A2+5] ^ B[A3+5];
-    Bp[A1+6] = B[A0+6] ^ B[A1+5] ^ B[A2+5] ^ B[A2+6] ^ B[A3+6];
-    Bp[A1+7] = B[A0+7] ^ B[A1+6] ^ B[A2+6] ^ B[A2+7] ^ B[A3+7];
-    
-
-    //  of = A2 ^ A3
-    //  A2 = A2 ^ (0x1b & ((signed char)of>>7));
-
-    //// A0 + A1
-    //  A2 = A2 ^ (A0)
-    //  A2 = A2 ^ (A1)
-
-    //// + 2 * A2
-    //  A2 = A2 ^ (A2 << 1)
-
-    //// + 3 * A3
-    //  A2 = A2 ^ (A3)
-    //  A2 = A2 ^ (A3<<1)
-
-
-    of = B[A2+7] ^ B[A3+7];
-
-    //          A0      A1          2*A2       2*A3         A3
-    Bp[A2+0] = B[A0+0] ^ B[A1+0]                     ^ B[A3+0] ^ of;
-    Bp[A2+1] = B[A0+1] ^ B[A1+1] ^ B[A2+0] ^ B[A3+0] ^ B[A3+1] ^ of;
-    Bp[A2+2] = B[A0+2] ^ B[A1+2] ^ B[A2+1] ^ B[A3+1] ^ B[A3+2];
-    Bp[A2+3] = B[A0+3] ^ B[A1+3] ^ B[A2+2] ^ B[A3+2] ^ B[A3+3] ^ of;
-    Bp[A2+4] = B[A0+4] ^ B[A1+4] ^ B[A2+3] ^ B[A3+3] ^ B[A3+4] ^ of;
-    Bp[A2+5] = B[A0+5] ^ B[A1+5] ^ B[A2+4] ^ B[A3+4] ^ B[A3+5];
-    Bp[A2+6] = B[A0+6] ^ B[A1+6] ^ B[A2+5] ^ B[A3+5] ^ B[A3+6];
-    Bp[A2+7] = B[A0+7] ^ B[A1+7] ^ B[A2+6] ^ B[A3+6] ^ B[A3+7];
-    
-
-    //  A3 = A0 ^ A3
-    //  A3 = A3 ^ (0x1b & ((signed char)of>>7));
-
-    //// 3 * A0
-    //  A3 = A3 ^ (A0)
-    //  A3 = A3 ^ (A0 << 1)
-
-    //// + A1 + A2
-    //  A3 = A3 ^ A1
-    //  A3 = A3 ^ A2
-
-    //// + 2 * A3
-    //  A3 = A3 ^ (A3<<1)
-
-    of = B[A0+7] ^ B[A3+7];
-
-    //        2*A0       A0         A1         A2       2*A3
-    Bp[A3+0] = B[A0+0] ^           B[A1+0] ^ B[A2+0]           ^ of;
-    Bp[A3+1] = B[A0+1] ^ B[A0+0] ^ B[A1+1] ^ B[A2+1] ^ B[A3+0] ^ of;
-    Bp[A3+2] = B[A0+2] ^ B[A0+1] ^ B[A1+2] ^ B[A2+2] ^ B[A3+1];
-    Bp[A3+3] = B[A0+3] ^ B[A0+2] ^ B[A1+3] ^ B[A2+3] ^ B[A3+2] ^ of;
-    Bp[A3+4] = B[A0+4] ^ B[A0+3] ^ B[A1+4] ^ B[A2+4] ^ B[A3+3] ^ of;
-    Bp[A3+5] = B[A0+5] ^ B[A0+4] ^ B[A1+5] ^ B[A2+5] ^ B[A3+4];
-    Bp[A3+6] = B[A0+6] ^ B[A0+5] ^ B[A1+6] ^ B[A2+6] ^ B[A3+5];
-    Bp[A3+7] = B[A0+7] ^ B[A0+6] ^ B[A1+7] ^ B[A2+7] ^ B[A3+6];
-    
-
-    //
-    Bp += BLOCK_SIZE/4;
-    B  += BLOCK_SIZE/4;
-    }
-
-
-    memmove(B - BLOCK_SIZE,Bp - BLOCK_SIZE,sizeof(Bp_space));
-}
-
-void bs_mixcolumns_rev(word_t * B)
-{
-    // to understand this, see
-    // https://en.wikipedia.org/wiki/Rijndael_mix_columns
-    // TODO combine with shiftrows for performance on decryption
-    word_t Bp_space[BLOCK_SIZE];
-    word_t * Bp = Bp_space;
-
- 
-    int i = 0;
-    for (; i < BLOCK_SIZE / 4; i += BLOCK_SIZE / 16)
-    {
-
-    //// state[i][0] = A0*0x0e + A1*0x0b + A2*0x0d + A3*0x09
-    // overflow:
-                    /* A0 * 0b1110             */   /* A1 * 0b1011     */   /* A2 * 0b1101   */   /* A3 * 0b1001 */
-    word_t of0 = (  (B[A0+7] ^ B[A0+6] ^ B[A0+5]) ^ (B[A1 + 7] ^ B[A1+5]) ^ (B[A2+6] ^ B[A2+5]) ^ ( B[A3+5] ));       // 2 bit
-    word_t of1 = (            (B[A0+7] ^ B[A0+6]) ^ (            B[A1+6]) ^ (B[A2+7] ^ B[A2+6]) ^ ( B[A3+6] ));       // 3 bit
-    word_t of2 = (                      (B[A0+7]) ^ (            B[A1+7]) ^ (          B[A2+7]) ^ ( B[A3+7] ));       // 4 bit
-
-    // inverse:
-    //          1110      1011    1101      1001
-    // A0 = A0 * 14 + A1 * 11 + A2 * 13 + A3 * 9
-    // A0 = A0 * (2+4+8) + A1 * (1+2+8) + A2 * (1+4+8) + A3 * (1+8)
-
-    //         (2*A0   +  4*A0   + 8*A0 )  + (8*A1   +  2*A1   +   A1  ) + ( A2    +   4*A2  +  8*A2 ) + (  A3   + 8*A3)
-    Bp[A0+0] =                                                   B[A1+0] ^ B[A2+0]                     ^ B[A3+0]           ^ of0;
-    Bp[A0+1] = B[A0+0]                               ^ B[A1+0] ^ B[A1+1] ^ B[A2+1]                     ^ B[A3+1]           ^ of0 ^ of1;
-    Bp[A0+2] = B[A0+1] ^ B[A0+0]                     ^ B[A1+1] ^ B[A1+2] ^ B[A2+2] ^ B[A2+0]           ^ B[A3+2]                 ^ of1 ^ of2;
-    Bp[A0+3] = B[A0+2] ^ B[A0+1] ^ B[A0+0] ^ B[A1+0] ^ B[A1+2] ^ B[A1+3] ^ B[A2+3] ^ B[A2+1] ^ B[A2+0] ^ B[A3+3] ^ B[A3+0] ^ of0       ^ of2;
-    Bp[A0+4] = B[A0+3] ^ B[A0+2] ^ B[A0+1] ^ B[A1+1] ^ B[A1+3] ^ B[A1+4] ^ B[A2+4] ^ B[A2+2] ^ B[A2+1] ^ B[A3+4] ^ B[A3+1] ^ of0 ^ of1;
-    Bp[A0+5] = B[A0+4] ^ B[A0+3] ^ B[A0+2] ^ B[A1+2] ^ B[A1+4] ^ B[A1+5] ^ B[A2+5] ^ B[A2+3] ^ B[A2+2] ^ B[A3+5] ^ B[A3+2]       ^ of1 ^ of2;
-    Bp[A0+6] = B[A0+5] ^ B[A0+4] ^ B[A0+3] ^ B[A1+3] ^ B[A1+5] ^ B[A1+6] ^ B[A2+6] ^ B[A2+4] ^ B[A2+3] ^ B[A3+6] ^ B[A3+3]             ^ of2;
-    Bp[A0+7] = B[A0+6] ^ B[A0+5] ^ B[A0+4] ^ B[A1+4] ^ B[A1+6] ^ B[A1+7] ^ B[A2+7] ^ B[A2+5] ^ B[A2+4] ^ B[A3+7] ^ B[A3+4];
-
-
-
-    //// state[i][1] = A0*0x09 + A1*0xe + A2*0x0b + A3*0x0d
-    // overflow:
-          /* A0 * 0b1001 */   /* A1 * 0b1110             */ /* A2 * 0b101  1     */   /* A3 * 0b1101 */
-    of0 = (  (B[A0+5]) ^      (B[A1+7] ^ B[A1+6] ^ B[A1+5]) ^ (B[A2 + 7] ^ B[A2+5]) ^ (B[A3+6] ^ B[A3+5]));       // 2 bit
-    of1 = (  (B[A0+6]) ^                (B[A1+7] ^ B[A1+6]) ^ (            B[A2+6]) ^ (B[A3+7] ^ B[A3+6]));       // 3 bit
-    of2 = (  (B[A0+7]) ^                          (B[A1+7]) ^ (            B[A2+7]) ^ (          B[A3+7]));       // 4 bit
-
-    // inverse:
-    //          1001      1110    1011      1101
-    // A1 = A0 * 9 + A1 * 14 + A2 * 11 + A3 * 13
-    // A1 = A0 * (1+8) + A1 * (2+4+8) + A2 * (1+2+8) + A3 * (1+4+8)
-
-    //         (1*A0   +  8*A0 ) +(2*A1    +  4*A1   +  8*A1 ) + (1*A2   +  2*A2   +  8*A2 ) + (1*A3   +   4*A3  +  8*A3)
-    Bp[A1+0] = B[A0+0] ^                                         B[A2+0] ^                     B[A3+0]                     ^ of0;
-    Bp[A1+1] = B[A0+1] ^           B[A1+0]                     ^ B[A2+1] ^ B[A2+0] ^           B[A3+1]                     ^ of0 ^ of1;
-    Bp[A1+2] = B[A0+2] ^           B[A1+1] ^ B[A1+0]           ^ B[A2+2] ^ B[A2+1] ^           B[A3+2] ^ B[A3+0]                 ^ of1 ^ of2;
-    Bp[A1+3] = B[A0+3] ^ B[A0+0] ^ B[A1+2] ^ B[A1+1] ^ B[A1+0] ^ B[A2+3] ^ B[A2+2] ^ B[A2+0] ^ B[A3+3] ^ B[A3+1] ^ B[A3+0] ^ of0       ^ of2;
-    Bp[A1+4] = B[A0+4] ^ B[A0+1] ^ B[A1+3] ^ B[A1+2] ^ B[A1+1] ^ B[A2+4] ^ B[A2+3] ^ B[A2+1] ^ B[A3+4] ^ B[A3+2] ^ B[A3+1] ^ of0 ^ of1;
-    Bp[A1+5] = B[A0+5] ^ B[A0+2] ^ B[A1+4] ^ B[A1+3] ^ B[A1+2] ^ B[A2+5] ^ B[A2+4] ^ B[A2+2] ^ B[A3+5] ^ B[A3+3] ^ B[A3+2]       ^ of1 ^ of2;
-    Bp[A1+6] = B[A0+6] ^ B[A0+3] ^ B[A1+5] ^ B[A1+4] ^ B[A1+3] ^ B[A2+6] ^ B[A2+5] ^ B[A2+3] ^ B[A3+6] ^ B[A3+4] ^ B[A3+3]             ^ of2;
-    Bp[A1+7] = B[A0+7] ^ B[A0+4] ^ B[A1+6] ^ B[A1+5] ^ B[A1+4] ^ B[A2+7] ^ B[A2+6] ^ B[A2+4] ^ B[A3+7] ^ B[A3+5] ^ B[A3+4];
-
-
-    //// state[i][2] = A0*0x0d + A1*0x09 + A2*0x0e + A3*0x0b
-    // overflow:
-          /* A1 * 0b1001 */  /* A2 * 0b1110              */ /* A3 * 0b1011      */   /* A0 * 0b1101 */
-    of0 = (  (B[A1+5]) ^      (B[A2+7] ^ B[A2+6] ^ B[A2+5]) ^ (B[A3 + 7] ^ B[A3+5]) ^ (B[A0+6] ^ B[A0+5]));       // 2 bit
-    of1 = (  (B[A1+6]) ^                (B[A2+7] ^ B[A2+6]) ^ (            B[A3+6]) ^ (B[A0+7] ^ B[A0+6]));       // 3 bit
-    of2 = (  (B[A1+7]) ^                          (B[A2+7]) ^ (            B[A3+7]) ^ (          B[A0+7]));       // 4 bit
-
-    // inverse:
-    //          1001      1110    1011      1101
-    // A2 = A1 * 9 + A2 * 14 + A3 * 11 + A0 * 13
-    // A2 = A1 * (1+8) + A2 * (2+4+8) + A3 * (1+2+8) + A0 * (1+4+8)
-
-    //         (1*A1   +  8*A1) + ( 2*A2   +  4*A2   +  8*A2)  + (1*A3      2*A2   +  8*A2) +  (1*A0   +   4*A0  +  8*A0)
-    Bp[A2+0] = B[A1+0] ^                                         B[A3+0] ^                     B[A0+0]                     ^ of0;
-    Bp[A2+1] = B[A1+1] ^           B[A2+0]                     ^ B[A3+1] ^ B[A3+0] ^           B[A0+1]                     ^ of0 ^ of1;
-    Bp[A2+2] = B[A1+2] ^           B[A2+1] ^ B[A2+0]           ^ B[A3+2] ^ B[A3+1] ^           B[A0+2] ^ B[A0+0]                 ^ of1 ^ of2;
-    Bp[A2+3] = B[A1+3] ^ B[A1+0] ^ B[A2+2] ^ B[A2+1] ^ B[A2+0] ^ B[A3+3] ^ B[A3+2] ^ B[A3+0] ^ B[A0+3] ^ B[A0+1] ^ B[A0+0] ^ of0       ^ of2;
-    Bp[A2+4] = B[A1+4] ^ B[A1+1] ^ B[A2+3] ^ B[A2+2] ^ B[A2+1] ^ B[A3+4] ^ B[A3+3] ^ B[A3+1] ^ B[A0+4] ^ B[A0+2] ^ B[A0+1] ^ of0 ^ of1;
-    Bp[A2+5] = B[A1+5] ^ B[A1+2] ^ B[A2+4] ^ B[A2+3] ^ B[A2+2] ^ B[A3+5] ^ B[A3+4] ^ B[A3+2] ^ B[A0+5] ^ B[A0+3] ^ B[A0+2]       ^ of1 ^ of2;
-    Bp[A2+6] = B[A1+6] ^ B[A1+3] ^ B[A2+5] ^ B[A2+4] ^ B[A2+3] ^ B[A3+6] ^ B[A3+5] ^ B[A3+3] ^ B[A0+6] ^ B[A0+4] ^ B[A0+3]             ^ of2;
-    Bp[A2+7] = B[A1+7] ^ B[A1+4] ^ B[A2+6] ^ B[A2+5] ^ B[A2+4] ^ B[A3+7] ^ B[A3+6] ^ B[A3+4] ^ B[A0+7] ^ B[A0+5] ^ B[A0+4];
-
-
-
-    //// state[i][3] = A0*0x0b + A1*0x0d + A2*0x09 + A3*0x0e
-    // overflow:
-           /* A2 * 0b1001 */  /* A3 * 0b1110             */ /* A0 * 0b1011     */     /* A1 * 0b1101 */
-    of0 = (  (B[A2+5]) ^      (B[A3+7] ^ B[A3+6] ^ B[A3+5]) ^ (B[A0 + 7] ^ B[A0+5]) ^ (B[A1+6] ^ B[A1+5]));       // 2 bit
-    of1 = (  (B[A2+6]) ^                (B[A3+7] ^ B[A3+6]) ^ (            B[A0+6]) ^ (B[A1+7] ^ B[A1+6]));       // 3 bit
-    of2 = (  (B[A2+7]) ^                          (B[A3+7]) ^ (            B[A0+7]) ^ (          B[A1+7]));       // 4 bit
-
-    // inverse:
-    //          1001      1110    1011      1101
-    // A2 = A2 * 9 + A3 * 14 + A0 * 11 + A1 * 13
-    // A2 = A2 * (1+8) + A3 * (2+4+8) + A0 * (1+2+8) + A1 * (1+4+8)
-
-    //         (1*A2   +  8*A2) + ( 2*A3   +  4*A3   +  8*A3)  + (1*A0      2*A0   +  8*A0) +  (1*A1   +   4*A1  +  8*A1)
-    Bp[A3+0] = B[A2+0] ^                                         B[A0+0] ^                     B[A1+0]                     ^ of0;
-    Bp[A3+1] = B[A2+1] ^           B[A3+0]                     ^ B[A0+1] ^ B[A0+0] ^           B[A1+1]                     ^ of0 ^ of1;
-    Bp[A3+2] = B[A2+2] ^           B[A3+1] ^ B[A3+0]           ^ B[A0+2] ^ B[A0+1] ^           B[A1+2] ^ B[A1+0]                 ^ of1 ^ of2;
-    Bp[A3+3] = B[A2+3] ^ B[A2+0] ^ B[A3+2] ^ B[A3+1] ^ B[A3+0] ^ B[A0+3] ^ B[A0+2] ^ B[A0+0] ^ B[A1+3] ^ B[A1+1] ^ B[A1+0] ^ of0       ^ of2;
-    Bp[A3+4] = B[A2+4] ^ B[A2+1] ^ B[A3+3] ^ B[A3+2] ^ B[A3+1] ^ B[A0+4] ^ B[A0+3] ^ B[A0+1] ^ B[A1+4] ^ B[A1+2] ^ B[A1+1] ^ of0 ^ of1;
-    Bp[A3+5] = B[A2+5] ^ B[A2+2] ^ B[A3+4] ^ B[A3+3] ^ B[A3+2] ^ B[A0+5] ^ B[A0+4] ^ B[A0+2] ^ B[A1+5] ^ B[A1+3] ^ B[A1+2]       ^ of1 ^ of2;
-    Bp[A3+6] = B[A2+6] ^ B[A2+3] ^ B[A3+5] ^ B[A3+4] ^ B[A3+3] ^ B[A0+6] ^ B[A0+5] ^ B[A0+3] ^ B[A1+6] ^ B[A1+4] ^ B[A1+3]             ^ of2;
-    Bp[A3+7] = B[A2+7] ^ B[A2+4] ^ B[A3+6] ^ B[A3+5] ^ B[A3+4] ^ B[A0+7] ^ B[A0+6] ^ B[A0+4] ^ B[A1+7] ^ B[A1+5] ^ B[A1+4];
-
-    Bp += BLOCK_SIZE/4;
-    B  += BLOCK_SIZE/4;
-    }
-
-    memmove(B - BLOCK_SIZE, Bp - BLOCK_SIZE,sizeof(Bp_space));
-
-}
-
-void bs_expand_key(word_t (* rk)[BLOCK_SIZE], uint8_t * _key)
-{
-    // TODO integrate this better
-    uint8_t key[KEY_SCHEDULE_SIZE];
-    memmove(key,_key,BLOCK_SIZE/8);
-    expand_key(key);
-
-    int i, j = 0, k, l;
-    for (i = 0; i < KEY_SCHEDULE_SIZE; i += (BLOCK_SIZE/8))
-    {
-        memmove(rk[j], key + i, BLOCK_SIZE / 8);
-
-        for (k = WORDS_PER_BLOCK; k < 128; k += WORDS_PER_BLOCK)
-        {
-            for (l = 0; l < WORDS_PER_BLOCK; l++)
-            {
-                rk[j][k + l] = rk[j][l];
-            }
-        }
-        bs_transpose(rk[j]);
-        j++;
-    }
-
-}
-
-void bs_cipher(word_t state[BLOCK_SIZE], word_t (* rk)[BLOCK_SIZE])
-{
-    int round;
-    bs_transpose(state);
-
-
-    bs_addroundkey(state,rk[0]);
-    for (round = 1; round < 10; round++)
-    {
-        bs_apply_sbox(state);
-        /*bs_shiftrows(state);*/
-        /*bs_mixcolumns(state);*/
-        bs_shiftmix(state);
-        bs_addroundkey(state,rk[round]);
-    }
-    bs_apply_sbox(state);
-    bs_shiftrows(state);
-    bs_addroundkey(state,rk[10]);
-    bs_transpose_rev(state);
-}
-
-void bs_cipher_rev(word_t state[BLOCK_SIZE], word_t (* rk)[BLOCK_SIZE])
-{
-    int round;
-    bs_transpose(state);
-
-    bs_addroundkey(state,rk[10]);
-    for (round = 9; round > 0; round--)
-    {
-        bs_shiftrows_rev(state);
-        bs_apply_sbox_rev(state);
-        bs_addroundkey(state,rk[round]);
-        bs_mixcolumns_rev(state);
-    }
-    bs_shiftrows_rev(state);
-    bs_apply_sbox_rev(state);
-    bs_addroundkey(state,rk[0]);
-
-    bs_transpose_rev(state);
-}
-
-void aes_ecb_test()
-{
-    uint8_t key_vector[16] = "\x2b\x7e\x15\x16\x28\xae\xd2\xa6\xab\xf7\x15\x88\x09\xcf\x4f\x3c";
-    uint8_t pt_vector[16] =  "\x6b\xc1\xbe\xe2\x2e\x40\x9f\x96\xe9\x3d\x7e\x11\x73\x93\x17\x2a";
-    uint8_t ct_vector[16] =  "\x3a\xd7\x7b\xb4\x0d\x7a\x36\x60\xa8\x9e\xca\xf3\x24\x66\xef\x97";
-    uint8_t output[16];
-    uint8_t input[16];
-    
-    printf("AES ECB\n");
-
-    aes_ecb_encrypt(output, pt_vector,16,key_vector);
-
-
-    printf("cipher text: \n");
-    dump_hex(output, 16);
-
-    aes_ecb_decrypt(input, output, 16, key_vector);
-
-    printf("plain text: \n");
-    dump_hex((uint8_t * )input,16);
-
-    if (memcmp(pt_vector, input, 16) != 0)
-    {
-        fprintf(stderr,"error: decrypted ciphertext is not the same as the input plaintext\n");
-        EXIT1;
-    }
-    else if (memcmp(ct_vector, output, 16) != 0)
-    {
-        fprintf(stderr,"error: ciphertext is not the same as the test vector\n");
-        EXIT1;
-    }
-    else
-    {
-        printf("ECB passes test vector\n\n");
-    }
-}
-
-void aes_ctr_test()
-{
-// Test vector from NIST for 4 input blocks
-#define AES_CTR_TESTS_BYTES 64
-
-    uint8_t key_vector[16] =
-        "\x2b\x7e\x15\x16\x28\xae\xd2\xa6\xab\xf7\x15\x88\x09\xcf\x4f\x3c";
-
-    uint8_t iv_vector[16]  =
-        "\xf0\xf1\xf2\xf3\xf4\xf5\xf6\xf7\xf8\xf9\xfa\xfb\xfc\xfd\xfe\xff";
-
-    uint8_t pt_vector[AES_CTR_TESTS_BYTES] =
-        "\x6b\xc1\xbe\xe2\x2e\x40\x9f\x96\xe9\x3d\x7e\x11\x73\x93\x17\x2a"
-        "\xae\x2d\x8a\x57\x1e\x03\xac\x9c\x9e\xb7\x6f\xac\x45\xaf\x8e\x51"
-        "\x30\xc8\x1c\x46\xa3\x5c\xe4\x11\xe5\xfb\xc1\x19\x1a\x0a\x52\xef"
-        "\xf6\x9f\x24\x45\xdf\x4f\x9b\x17\xad\x2b\x41\x7b\xe6\x6c\x37\x10"
-        ;
-
-    uint8_t ct_vector[AES_CTR_TESTS_BYTES] =
-        "\x87\x4d\x61\x91\xb6\x20\xe3\x26\x1b\xef\x68\x64\x99\x0d\xb6\xce"
-        "\x98\x06\xf6\x6b\x79\x70\xfd\xff\x86\x17\x18\x7b\xb9\xff\xfd\xff"
-        "\x5a\xe4\xdf\x3e\xdb\xd5\xd3\x5e\x5b\x4f\x09\x02\x0d\xb0\x3e\xab"
-        "\x1e\x03\x1d\xda\x2f\xbe\x03\xd1\x79\x21\x70\xa0\xf3\x00\x9c\xee"
-        ;
-
-    uint8_t output[AES_CTR_TESTS_BYTES];
-    uint8_t input[AES_CTR_TESTS_BYTES];
-
-    printf("AES CTR\n");
-
-    aes_ctr_encrypt(output,pt_vector,AES_CTR_TESTS_BYTES,key_vector, iv_vector);
-
-    printf("cipher text: \n");
-    dump_hex(output,AES_CTR_TESTS_BYTES);
-    
-    aes_ctr_decrypt(input,output,AES_CTR_TESTS_BYTES,key_vector, iv_vector);
-
-    printf("plain text: \n");
-    dump_hex(input,AES_CTR_TESTS_BYTES);
-					  
-    if (memcmp(pt_vector, input, AES_CTR_TESTS_BYTES) != 0)
-    {
-        fprintf(stderr,"error: decrypted ciphertext is not the same as the input plaintext\n");
-        EXIT1;
-    }
-    else if (memcmp(ct_vector, output, AES_CTR_TESTS_BYTES) != 0)
-    {
-        fprintf(stderr,"error: ciphertext is not the same as the test vector\n");
-        EXIT1;
-    }
-    else
-    {
-        printf("CTR passes test vector\n\n");
-    }
-
+int i[1];
+int j, bs_transpose_dst_k, k, s, o;
+void a(int (*)[], uint8_t *);
+void b(uint8_t c, uint8_t d, size_t e, uint8_t f, uint8_t g) {
+  int l[1];
+  a(l, f);
+}
+void a(int (*l)[], uint8_t *m) {
+  for (; o < 76; o += 8) {
+    {
+      int *n = i;
+      bs_transpose_dst_k = 0;
+      for (; bs_transpose_dst_k < 64; bs_transpose_dst_k++) {
+        j = 0;
+        for (; j < 64; j++) {
+          k = &s;
+          n[j] = k & 1;
+        }
+      }
+    }
+  }
+}
+void aes_ecb_test() {}
+void aes_ctr_test() {
+  uint8_t p = "";
+  uint8_t q = "";
+  uint8_t r = "";
+  uint8_t output[4];
+  b(output, r, 4, p, q);
 }
diff --git a/test/monniaux/bitsliced-aes/one_file/reduce/bitsliced-aes_main.c b/test/monniaux/bitsliced-aes/one_file/reduce/bitsliced-aes_main.c
index 71f67f0f..0d48b3b8 100644
--- a/test/monniaux/bitsliced-aes/one_file/reduce/bitsliced-aes_main.c
+++ b/test/monniaux/bitsliced-aes/one_file/reduce/bitsliced-aes_main.c
@@ -1,4 +1,4 @@
-#include "../../../clock.h"
+#include "/home/monniaux/work/Kalray/CompCert/test/monniaux/clock.h"
 
 void aes_ecb_test(void);
 void aes_ctr_test(void);
diff --git a/test/monniaux/bitsliced-aes/one_file/reduce/compare.sh b/test/monniaux/bitsliced-aes/one_file/reduce/compare.sh
index e98f4677..a21bb465 100755
--- a/test/monniaux/bitsliced-aes/one_file/reduce/compare.sh
+++ b/test/monniaux/bitsliced-aes/one_file/reduce/compare.sh
@@ -1,7 +1,7 @@
 #!/bin/bash
 ROOT=/home/monniaux/work/Kalray/CompCert
 SRC=bitsliced-aes_compute.c
-MAIN=bitsliced-aes_main
+MAIN=/home/monniaux/work/Kalray/CompCert/test/monniaux/bitsliced-aes/one_file/reduce/bitsliced-aes_main
 k1-mbr-gcc -Werror=implicit -Werror=uninitialized -O3 $SRC $ROOT/test/monniaux/clock.gcc.k1c.o $MAIN.gcc.k1c.o -o bitsliced-aes.gcc.k1c &&
 $ROOT/ccomp -O3 -fno-unprototyped -O3 $SRC $ROOT/test/monniaux/clock.gcc.k1c.o  $MAIN.gcc.k1c.o -o bitsliced-aes.ccomp.k1c &&
 gcc -Werror=implicit -Werror=uninitialized -O3 $SRC  $ROOT/test/monniaux/clock.gcc.host.o $MAIN.c -o bitsliced-aes.gcc.host &&
@@ -12,5 +12,5 @@ grep cycles ./bitsliced-aes.gcc.k1c.out > ./bitsliced-aes.gcc.k1c.cycles &&
 grep cycles ./bitsliced-aes.ccomp.k1c.out > ./bitsliced-aes.ccomp.k1c.cycles &&
 sed -i -e 's/cycles: //' ./bitsliced-aes.gcc.k1c.cycles &&
 sed -i -e 's/cycles: //' ./bitsliced-aes.ccomp.k1c.cycles &&
-test $(cat ./bitsliced-aes.ccomp.k1c.cycles) -gt 300 &&
+test $(cat ./bitsliced-aes.gcc.k1c.cycles) -gt 100000 &&
 test $(cat ./bitsliced-aes.ccomp.k1c.cycles) -gt $(expr 2 '*' $(cat ./bitsliced-aes.gcc.k1c.cycles))