flint3-sys 3.2.2-1

Rust bindings to the FLINT C library
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135

/*
    Copyright (C) 2010 Sebastian Pancratz
    Copyright (C) 2011 William Hart

    This file is part of FLINT.

    FLINT is free software: you can redistribute it and/or modify it under
    the terms of the GNU Lesser General Public License (LGPL) as published
    by the Free Software Foundation; either version 3 of the License, or
    (at your option) any later version.  See <https://www.gnu.org/licenses/>.
*/

#include "nmod.h"
#include "nmod_poly.h"

void _nmod_poly_rem_q1(nn_ptr R,
                       nn_srcptr A, slong lenA, nn_srcptr B, slong lenB,
                       nmod_t mod)
{
    slong i;
    ulong invL, t, q0, q1, t2, t1, t0, s1, s0;

    FLINT_ASSERT(lenA == lenB + 1);
    invL = (B[lenB-1] == 1) ? 1 : n_invmod(B[lenB-1], mod.n);

    if (lenB < 2)
        return;

    q1 = n_mulmod2_preinv(A[lenA-1], invL, mod.n, mod.ninv);
    t  = n_mulmod2_preinv(q1, B[lenB-2], mod.n, mod.ninv);
    t  = n_submod(t, A[lenA-2], mod.n);
    q0 = n_mulmod2_preinv(t, invL, mod.n, mod.ninv);
    q1 = nmod_neg(q1, mod);

    /* R = A + (q1*x + q0)*B */
    t = A[0];
    NMOD_ADDMUL(t, q0, B[0], mod);
    R[0] = t;

    if (mod.norm >= (FLINT_BITS + 1)/2 + 1)
    {
        for (i = 1; i < lenB - 1; i++)
        {
            NMOD_RED2(R[i], 0, A[i] + q1*B[i - 1] + q0*B[i], mod);
        }
    }
    else if (mod.norm != 0)
    {
        for (i = 1; i < lenB - 1; i++)
        {
            umul_ppmm(t1, t0, q1, B[i - 1]);
            umul_ppmm(s1, s0, q0, B[i]);
            add_ssaaaa(t1, t0, t1, t0, 0, A[i]);
            add_ssaaaa(t1, t0, t1, t0, s1, s0);
            t1 = FLINT_MIN(t1, t1 - mod.n);
            FLINT_ASSERT(t1 < mod.n);
            NMOD_RED2(R[i], t1, t0, mod);
        }
    }
    else
    {
        for (i = 1; i < lenB - 1; i++)
        {
            umul_ppmm(t1, t0, q1, B[i - 1]);
            umul_ppmm(s1, s0, q0, B[i]);
            add_ssaaaa(t1, t0, t1, t0, 0, A[i]);
            add_sssaaaaaa(t2, t1, t0, 0, t1, t0, 0, s1, s0);
            if (t2 != 0)
                /* Note: should just be t1 -= mod.n, but with GCC
                   on Zen3 that version runs noticeably slower. */
                sub_ddmmss(t2, t1, t2, t1, 0, mod.n);
            t1 = FLINT_MIN(t1, t1 - mod.n);
            FLINT_ASSERT(t1 < mod.n);
            NMOD_RED2(R[i], t1, t0, mod);
        }
    }
}

void _nmod_poly_rem(nn_ptr R, nn_srcptr A, slong lenA,
                              nn_srcptr B, slong lenB, nmod_t mod)
{
    if (lenA - lenB == 1)
    {
        _nmod_poly_rem_q1(R, A, lenA, B, lenB, mod);
    }
    else if (lenB >= 2)
    {
        nn_ptr Q;
        TMP_INIT;

        TMP_START;
        Q = TMP_ALLOC((lenA - lenB + 1) * sizeof(ulong));
        _nmod_poly_divrem(Q, R, A, lenA, B, lenB, mod);
        TMP_END;
    }
}

void nmod_poly_rem(nmod_poly_t R, const nmod_poly_t A, const nmod_poly_t B)
{
    const slong lenA = A->length, lenB = B->length;
    nmod_poly_t tR;
    nn_ptr r;

    if (lenB == 0)
    {
        flint_throw(FLINT_ERROR, "Exception (nmod_poly_rem). Division by zero.\n");
    }
    if (lenA < lenB)
    {
        nmod_poly_set(R, A);
        return;
    }

    if (R == B)
    {
        nmod_poly_init2_preinv(tR, B->mod.n, B->mod.ninv, lenB - 1);
        r = tR->coeffs;
    }
    else
    {
        nmod_poly_fit_length(R, lenB - 1);
        r = R->coeffs;
    }

    _nmod_poly_rem(r, A->coeffs, lenA, B->coeffs, lenB, A->mod);

    if (R == B)
    {
        nmod_poly_swap(R, tR);
        nmod_poly_clear(tR);
    }

    R->length = lenB - 1;
    _nmod_poly_normalise(R);
}