g1p_sub.cu

Go to the documentation of this file.

// bls12_381: Arithmetic for BLS12-381
// Copyright 2022-2023 Dag Arne Osvik
// Copyright 2022-2023 Luan Cardoso dos Santos
 
// #include <stdio.h>
 
#include "fp.cuh"
#include "g1.cuh"
 
__device__ void g1p_sub(g1p_t &p, const g1p_t &q) {
 
    fp_t
        X1, Y1, Z1,
        X2, Y2, Z2,
        t0, t1, t2, t3;
 
    fp_cpy(X1, p.x);
    fp_cpy(Y1, p.y);
    fp_cpy(Z1, p.z);
 
    fp_cpy(X2, q.x);
    fp_cpy(Y2, q.y);
    fp_cpy(Z2, q.z);
 
    if (!g1p_isInf(q))
        fp_neg(Y2, Y2);
 
    // Adapted from eprint 2015-1060, algorithm 7.
    // Modified to avoid overwriting inputs and remove one temp value.
    // 12 mul, 0 square, 11 add, 5 sub, 2 x12, 1 x3.
 
    fp_add(t0, X1, Y1); // t3
 
    //fp_print("T3 = ",  t0);
 
    fp_add(t1, Y1, Z1); // t8
 
    //fp_print("T8 = ",  t1);
 
    fp_add(t2, Z1, X1); // td
 
    //fp_print("Td = ",  t2);
 
 
    fp_mul(X1, X1, X2); // t0
 
    //fp_print("T0 = ",  X1);
 
    fp_mul(Y1, Y1, Y2); // t1
 
    //fp_print("T1 = ",  Y1);
 
    fp_mul(Z1, Z1, Z2); // t2
 
    //fp_print("T2 = ",  Z1);
 
 
    fp_add(t3, X2, Y2); // t4
 
    //fp_print("T4 = ",  t3);
 
    fp_add(Y2, Y2, Z2); // t9
 
    //fp_print("T9 = ",  Y2);
 
    fp_add(Z2, Z2, X2); // te
 
    //fp_print("Te = ",  Z2);
 
 
    fp_mul(X2, t3, t0); // t5
 
    //fp_print("T5 = ",  X2);
 
    fp_mul(Y2, Y2, t1); // ta
 
    //fp_print("Ta = ",  Y2);
 
    fp_mul(Z2, Z2, t2); // tf
 
    //fp_print("Tf = ",  Z2);
 
 
    fp_x3(t0, X1);      // ti
 
    //fp_print("Ti = ",  t0);
 
    fp_add(t1, Y1, Z1); // tb
 
    //fp_print("Tb = ",  t1);
 
    fp_add(t2, Z1, X1); // tg
 
    //fp_print("Tg = ",  t2);
 
    fp_x12(t3, Z1);     // tk
 
    //fp_print("Tk = ",  t3);
 
 
    fp_add(X1, X1, Y1); // t6
 
    //fp_print("T6 = ",  X1);
 
    fp_add(Z1, Y1, t3); // tl
 
    //fp_print("Tl = ",  Z1);
 
    fp_sub(Y1, Y1, t3); // tm
 
    //fp_print("Tm = ",  Y1);
 
 
    fp_sub(X1, X2, X1); // t7
 
    //fp_print("T7 = ",  X1);
 
    fp_mul(X2, X1, t0); // ts
 
    //fp_print("Ts = ",  X2);
 
 
    fp_mul(X1, X1, Y1); // tp
 
    //fp_print("Tp = ",  X1);
 
    fp_mul(Y1, Y1, Z1); // tr
 
    //fp_print("Tr = ",  Y1);
 
 
    fp_sub(Y2, Y2, t1); // tc
 
    //fp_print("Tc = ",  Y2);
 
    fp_mul(Z1, Z1, Y2); // tt
 
    //fp_print("Tt = ",  Z1);
 
    fp_sub(Z2, Z2, t2); // th
 
    //fp_print("Th = ",  Z2);
 
 
    fp_x12(Z2, Z2);     // tn
 
    //fp_print("Tn = ",  Z2);
 
    fp_mul(Y2, Y2, Z2); // to
 
    //fp_print("To = ",  Y2);
 
    fp_mul(Z2, Z2, t0); // tq
 
    //fp_print("Tq = ",  Z2);
 
 
    fp_sub(X1, X1, Y2); // X3
    fp_add(Y1, Y1, Z2); // Y3
    fp_add(Z1, Z1, X2); // Z3
 
    //fp_print("X4 = ",  X1);
    //fp_print("Y4 = ",  Y1);
    //fp_print("Z4 = ",  Z1);
 
    fp_cpy(p.x, X1);
    fp_cpy(p.y, Y1);
    fp_cpy(p.z, Z1);
}
 
// vim: ts=4 et sw=4 si