isogeny_walks_3_fp.c File Reference

#include <stdio.h>
#include "isogeny_walks.h"
#include "utilities.h"

Include dependency graph for isogeny_walks_3_fp.c:

Go to the source code of this file.

Macros
#define	TOP_BIT   (8*sizeof(int)-1)

Functions
void	isogeny_walks_get_points_3_fp (fp_t *output, const fp_t *input_A)
void	isogeny_walks_to_montgomery_model_3_fp (fp_t *output_A, const fp_t *input_t)
void	isogeny_to_montgomery (fp_t output_A, const fp_t *input_t)
void	isogeny_walks_3_fp (fp_t output_A, const fp_t input_A, int input_length)

Macro Definition Documentation

TOP_BIT

#define TOP_BIT   (8*sizeof(int)-1)

Definition at line 210 of file isogeny_walks_3_fp.c.

Referenced by isogeny_walks_3_fp().

Function Documentation

isogeny_to_montgomery()

void isogeny_to_montgomery	(	fp_t	output_A,
		const fp_t *	input_t )

Definition at line 183 of file isogeny_walks_3_fp.c.

                                                              {
    fp_t r_squared, r_at_four, s_squared, s_at_four;
    fp_t num, den, aux;
 
    // input_t[0] is r
    // input_t[1] is s
    fp_sqr(r_squared, input_t[0]);
    fp_sqr(s_squared, input_t[1]);
    fp_sqr(r_at_four, r_squared);
    fp_sqr(s_at_four, s_squared);
 
    // num = -3 * r_at_four - 6 * r_squared * s_squared + s_at_four
    fp_add(aux, r_squared, r_squared);
    fp_mul(num, aux, s_squared);
    fp_add(num, num, r_at_four);
    fp_sub(s_at_four, s_at_four, num);
    fp_sub(s_at_four, s_at_four, num);
    fp_sub(num, s_at_four, num);
 
    // den = 4 * r_squared * r * s
    fp_add(aux, aux, aux);
    fp_mul(den, aux, input_t[0]);
    fp_mul(den, den, input_t[1]);
    fp_inv(den, den);
    fp_mul(output_A, den, num);
}

References fp_add, fp_inv, fp_mul, fp_sqr, fp_sub, and num.

Referenced by isogeny_walks_3_fp(), and isogeny_walks_3_fp().

Here is the caller graph for this function:

isogeny_walks_3_fp()

void isogeny_walks_3_fp	(	fp_t	output_A,
		const fp_t	input_A,
		int	input_length )

Definition at line 217 of file isogeny_walks_3_fp.c.

                                                                            {
    fp_t aux[2], num_den[3];
 
    fp_copy(aux[0], input_A);
    fp_set_to_one(aux[1]);
 
    isogeny_walks_get_points_3_fp(num_den, aux);
 
    uint64_t sign = (input_length >> TOP_BIT) & 1;
    fp_cset(aux[0], num_den[0], -sign);
    fp_cset(aux[0], num_den[1], sign-1);
    fp_copy(aux[1], num_den[2]);
 
    int input_length_abs = absolute_value(input_length);
    for(int i = 0; i < input_length_abs; i++){
        fp_t t[2] = {0};
        uint64_t d = constant_time_is_lessthan_u64(i, input_length_abs);
        isogeny_walks_to_montgomery_model_3_fp(t, aux);
        fp_cset(aux[0], t[0], -d);
        fp_cset(aux[1], t[1], -d);
    }
    isogeny_to_montgomery(output_A, aux);
}

References constant_time_is_lessthan_u64(), fp_copy, fp_cset(), fp_set_to_one(), i, isogeny_to_montgomery(), isogeny_walks_get_points_3_fp(), isogeny_walks_to_montgomery_model_3_fp(), and TOP_BIT.

Referenced by action().

Here is the call graph for this function:

Here is the caller graph for this function:

isogeny_walks_get_points_3_fp()

void isogeny_walks_get_points_3_fp	(	fp_t *	output,
		const fp_t *	input_A )

Definition at line 18 of file isogeny_walks_3_fp.c.

                                                                     {
    fp_t A_times_one_third, A_squared, C_squared;
    fp_t t, r, s, u, aux, tmp, y;
    fp_t s0, s1, s2, s0_squared, v;
 
    // input_A[0] is A
    // input_A[1] is C
    // A_times_one_third = A * one_third                    #               M𝒸
    fp_mul(A_times_one_third, ONE_THIRD, input_A[0]);
    // A_squared = A**2                                     #               S
    fp_sqr(A_squared, input_A[0]);
    // C_squared = C**2                                     #               S
    fp_sqr(C_squared, input_A[1]);
 
    // t = (3 * C_squared - A_squared) * one_ninth          #               M𝒸 + 3A
    fp_add(t, C_squared, C_squared);
    fp_add(t, t, C_squared);
    fp_sub(t, t, A_squared);
    fp_mul(t, ONE_NINTH, t);
 
    // r = 16 * (A_squared * A) * one_by_27                 #               M + M𝒸 + 4A
    fp_mul(r, A_squared, input_A[0]);
    fp_mul(r, ONE_BY_27, r);
    fp_add(r, r, r);
    fp_add(r, r, r);
    fp_add(r, r, r);
    fp_add(r, r, r);
 
    // s = 8 * A_times_one_third                            #                 3A
    fp_add(s, A_times_one_third, A_times_one_third);
    fp_add(s, s, s);
    fp_add(s, s, s);
 
    // u = r - s * C_squared                                #                M  +  A
    fp_mul(u, s, C_squared);
    fp_sub(u, r, u);
 
    // aux = 2*C_squared**2                                         #               S + A
    fp_sqr(aux, C_squared);
    fp_add(aux, aux, aux);
    // y = 4 * C_squared                                            #               2A
    fp_add(y, C_squared, C_squared);
    fp_add(y, y, y);
    // y = y - A_squared                                            #               A
    fp_sub(y, y, A_squared);
    // y = aux * y                                                  #               M
    fp_mul(y, aux, y);
 
    // y = t + one_third * setup['curt'](y)                         #            M𝒸 + A + E   (TOTAL = S + M + Mc + 5A + E)
    fp_curt(y, y);
    fp_mul(y, y, ONE_THIRD);
    fp_add(y, y, t);
 
    // r = 2 * y                                                    #                  A
    fp_add(r, y, y);
    // s = 6 * t                                                    #                 3A
    fp_add(s, t, t);
    fp_add(t, s, s);
    fp_add(s, s, t);
 
    // s0 = setup['sqrt'](r - s)                                    #                  A + E
    fp_sub(s0, r, s);
    fp_sqrt(s0, s0);
 
    // s0_squared = s0**2                                           #                 S
    fp_sqr(s0_squared, s0);
    // v = -(r + s)                                                 #                 2A
    fp_add(v, r, s);
    fp_neg(v, v);
 
    fp_mul(tmp, v, s0_squared); //               #                  M
    fp_mul(u, u, s0);           //               #                  M
    // s1 = setup['sqrt'](tmp + u)                                  #                A + E
    fp_add(t, tmp, u);
    fp_sqrt(s1, t);
    // s2 = setup['sqrt'](tmp - u)                                  #                A + E
    fp_sub(s, tmp, u);
    fp_sqrt(s2, s);
 
    // Test to know if s1 and s2 are square roots                   #                 2S
    fp_sqr(tmp, s1);
    fp_sqr(aux, s2);
    uint8_t test_s1 = fp_is_equal(tmp, t);
    uint8_t test_s2 = fp_is_equal(aux, s);
 
    // Compute z                                                    #                 4Mc + 5A
    fp_sub(t, s1, s0_squared);
    fp_mul(t, t, ONE_HALF);
    fp_neg(s1, s1);
    fp_sub(r, s1, s0_squared);
    fp_mul(r, r, ONE_HALF);
 
    fp_add(s, s2, s0_squared);
    fp_mul(s, s, ONE_HALF);
    fp_sub(u, s0_squared, s2);
    fp_mul(u, u, ONE_HALF);
 
    // Select z in constant time
    fp_cset(output[0], t, -test_s1);
    fp_cset(output[1], r, -test_s1);
    fp_cset(output[0], s, -test_s2);
    fp_cset(output[1], u, -test_s2);
 
    // Compute num and den                                          #                  2M + 2A
    fp_mul(output[2], s0, input_A[1]);
    fp_mul(aux, s0, A_times_one_third);
    fp_sub(output[0], output[0], aux);
    fp_sub(output[1], output[1], aux);
 
    //                                                              #                 2S + 7M + 2A
    fp_sqr(r, output[0]); // x**2
    fp_sqr(s, output[2]); // den**2
    fp_mul(v, input_A[1], output[0]); // C*x
    fp_mul(t, output[2], input_A[0]); // A*den
    fp_mul(u, v, s); // C*x*den**2
    fp_mul(aux, s, output[2]); // den**3
    fp_mul(aux, aux, input_A[1]); // C*den**3
    fp_add(t, t, v); // C*x + A*den
    fp_mul(t, t, r); // x**2 * (C*x + A*den)
    fp_add(t, t, u); // (x**2 * (C*x + A*den)) + C*x * den**2
    fp_mul(t, t, aux);
 
    uint64_t test_sqr = (uint64_t)!fp_is_square(t);//          #                     E
    fp_cswap(output[0], output[1], -test_sqr);
}

References fp_add, fp_cset(), fp_cswap, fp_curt(), fp_is_equal(), fp_is_square(), fp_mul, fp_neg(), fp_sqr, fp_sqrt(), and fp_sub.

Referenced by isogeny_walks_3_fp(), and isogeny_walks_3_fp().

Here is the call graph for this function:

Here is the caller graph for this function:

isogeny_walks_to_montgomery_model_3_fp()

void isogeny_walks_to_montgomery_model_3_fp	(	fp_t *	output_A,
		const fp_t *	input_t )

Definition at line 144 of file isogeny_walks_3_fp.c.

                                                                                {
    fp_t r_squared, s_squared, r_cubed;
    fp_t aux, alpha_cubed, alpha, rd;
 
    // input_t[0] is r
    // input_t[1] is s
    // r_squared = r**2
    fp_sqr(r_squared, input_t[0]);
    // s_squared = s**2
    fp_sqr(s_squared, input_t[1]);
    // r_cubed = r_squared * r
    fp_mul(r_cubed, r_squared, input_t[0]);
    // s_cubed = s_squared * s
    fp_mul(output_A[1], s_squared, input_t[1]);
    // aux = r * s_squared
    fp_mul(aux, s_squared, input_t[0]);
 
    // alpha_cubed = r * (r_squared - s_squared)
    fp_sub(alpha_cubed, r_squared, s_squared);
    fp_mul(alpha_cubed, alpha_cubed, input_t[0]);
    // alpha = setup['curt'](alpha_cubed)
    fp_curt(alpha, alpha_cubed);
 
    // Goal rd = 3 * (r * alpha**2 + r_squared * alpha + r_cubed) - s_sqr * alpha - 2 * aux
    //rd = (r * alpha**2 + r_squared * alpha + r_cubed)
    fp_mul(rd, input_t[0], alpha);
    fp_add(rd, rd, r_squared);
    fp_mul(rd, rd, alpha);
    fp_add(rd, rd, r_cubed);
    // rd = 3 * (r * alpha**2 + r_squared * alpha + r_cubed)
    fp_add(r_squared, rd, rd);
    fp_add(rd, rd, r_squared);
    // rd = 3 * (r * alpha**2 + r_squared * alpha + r_cubed) - s_sqr * alpha - 2 * aux
    fp_sub(rd, rd, aux);
    fp_sub(rd, rd, aux);
    fp_mul(r_squared, s_squared, alpha);
    fp_sub(output_A[0], rd, r_squared);
}

References fp_add, fp_curt(), fp_mul, fp_sqr, and fp_sub.

Referenced by isogeny_walks_3_fp(), and isogeny_walks_3_fp().

Here is the call graph for this function:

Here is the caller graph for this function: