Quick implementation of Galois fields

galois.c

/*
 * The following is an implementation of the finite field GF(2^8) as bit vectors of length 8, where the nth bit represents the
 * coefficient of the nth power of the generator in each element, and the generator satisfies the minimal polynomial
 * x^8 + x^4 + x ^3 + x^2 + 1 in the prime field Z_2, in which addition is equivalent to XOR and multiplication to AND.
 * The elements of GF(2^8) thus represent polynomials of degree < 8 in the generator x. Addition in this field is simply 
 * bitwise XOR, but multiplication requires the elimination of powers of x <= 8.
 */

#include <stdio.h>
#include <stdint.h>

typedef uint8_t gal8; /* Galois field of order 2^8 */

const gal8 min_poly  = 0b11101,     /* Minimal polynomial x^8 + x^4 + x^3 + x^2 + 1 */
           generator = 0b10;        /* Generator of Galois field */

gal8 gal_add(gal8 a, gal8 b);       /* Add two elements of GF(2^8) */
gal8 gal_mul(gal8 a, gal8 b);       /* Multiply two elements of GF(2^8) */
void gal_print(gal8 a);             /* Print an element of GF(2^8) in binary form */
int hamming_norm(int a);            /* Number of nonzero bits in a */
int hamming_distance(int a, int b); /* Number of different bits between a and b */

int main()
{
    int i = 0, c = 0;
    gal8 a = 1;
    while (i++ < 256) {
        gal_print(a);
        /* printf("%d", hamming_distance(a, gal_mul(a, generator))); */
        a = gal_mul(a, generator);
        if (c++ < 7) {
            putchar(' ');
        } else {
            putchar('\n');
            c = 0;
        }
    }
    if (c)
        putchar('\n');
}

gal8 gal_add(gal8 a, gal8 b)
{
    return a ^ b;
}

gal8 gal_mul(gal8 a, gal8 b)
{
    gal8 res = 0;
    for (; b; b >>= 1) {
        if (b & 1)
            res ^= a;
        if (a & 0x80)
            a = (a << 1) ^ min_poly;
        else
            a <<= 1;
    }
    return res;
}

void gal_print(gal8 a)
{
    int i = 8;
    while (i--)
        putchar((a >> i & 1) + '0');
}

int hamming_norm(int a)
{
    int res = 0;
    while (a) {
        if (a & 1) ++res;
        a >>= 1;
    }
    return res;
}

int hamming_distance(int a, int b)
{
    return hamming_norm(a ^ b);
}

galois_matrix.c

/* 
 * This alternate implementation realizes elements of the finite field GF(2^8) as 8x8 matrices of bits, or arrays of 8 bytes.
 * The generator of the finite field is simply the companion matrix of the minimal polynomial x^8 + x^4 + x^3 + x^2 + 1,
 * or its transpose. The arithmetic in this finite field is usual matrix arithmetic.
 */

#include <stdio.h>
#include <stdint.h>
#include <unistd.h> /* for sleep */

typedef uint8_t gal8;
             
const gal8 gener[8] = {1, 0x80, 0x40 + 1, 0x20 + 1, 0x10 + 1, 8, 4, 2}, /* companion matrix of min_poly */
           gentr[8] = {0x40, 0x20, 0x10, 8, 4, 2, 1, 0b10111000}, /* transpose of gener */
           unity[8] = {0x80, 0x40, 0x20, 0x10, 8, 4, 2, 1},
           zero[8]  = {0};

gal8 *mat_add(gal8 *a, gal8 *b);          /* add two matrices */
gal8 *mat_mul(gal8 *a, gal8 *b);          /* multiply two matrices */
gal8 *mat_scalar_mul(uint8_t a, gal8 *b); /* multiply matrix by scalar */
gal8 mat_app(gal8 a, gal8 *b);            /* apply matrix to row vector */
gal8 *to_matrix(gal8 a);                  /* convert byte into matrix */
void mat_print(gal8 *a);                  /* print matrix */

int main()
{
    int i, j;
    gal8 a[8] = {0}, *aux;
    
    /* copy unity onto a */
    for (i = 0; i < 8; ++i)
        a[i] = unity[i];
    
    /* print each power of gentr */
    for (j = 0; j < 256; ++j) {
        mat_print(a);
        aux = mat_mul(a, gentr);
        for (i = 0; i < 8; ++i)
            a[i] = aux[i];
        sleep(1);
        putchar('\n');
    }
}

gal8 *mat_add(gal8 *a, gal8 *b)
{
    gal8 res[8];
    int i;
    
    for (i = 0; i < 8; ++i)
        res[i] = a[i] + b[i];
        
    return res;
}

gal8 *mat_mul(gal8 *a, gal8 *b)
{
    gal8 res[8] = {0};
    int i, j, k;
    
    /* For each ij, res_ij = sum of a_ik & b_kj for each k */
    for (i = 0; i < 8; ++i)
        for (j = 0; j < 8; ++j)
            for (k = 0; k < 8; ++k)
                res[i] ^= ((a[i] >> (7 - k)) & (b[k] >> (7 - j)) & 1) << (7 - j); 
    
    return res;
}

/* this concerns a as a scalar, not as a gal8 */
gal8 *mat_scalar_mul(uint8_t a, gal8 *b)
{
    gal8 res[8];
    int i;
    
    for (i = 0; i < 8; ++i)
        res[i] = a & b[i];
    
    return res;
}

/* a is taken to be a row vector, applied to b from the left */
gal8 mat_app(gal8 a, gal8 *b)
{
    gal8 res = 0;
    int i, j;
    
    /* multiply the ith row of b with the ith bit of a starting from the left */
    for (i = 0, j = 0x80; i < 8; ++i, j >>= 1)
        if (a & j)
            res ^= b[i];
    
    return res;
}

/* convert a to its matrix equivalent */
gal8 *to_matrix(gal8 a)
{
    gal8 res[8] = {0}, *aux;
    int i;
    
    for (; a; a >>= 1) {
        aux = mat_add(res, mat_scalar_mul(a & 1, gener));
        for (i = 0; i < 8; ++i)
            res[i] = aux[i];
    }

    return res;
}

void mat_print(gal8 *a)
{
    int i;
    for (i = 0; i < 8; ++i) {
        gal_print(a[i]);
        putchar('\n');
    }
}

First prog is ok.But 2nd prog i.e with matrix is giving errors.pl see

galois.c line 14 IS WRONG, use 0b00011011 instead

For posterity's sake, the gal_mul function uses a modified version of the peasant's algorithm. The algorithm with explanation itself is described on Wikipedia.