1 /*
2 * Copyright (c) 2008-2020 Stefan Krah. All rights reserved.
3 *
4 * Redistribution and use in source and binary forms, with or without
5 * modification, are permitted provided that the following conditions
6 * are met:
7 *
8 * 1. Redistributions of source code must retain the above copyright
9 * notice, this list of conditions and the following disclaimer.
10 *
11 * 2. Redistributions in binary form must reproduce the above copyright
12 * notice, this list of conditions and the following disclaimer in the
13 * documentation and/or other materials provided with the distribution.
14 *
15 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS "AS IS" AND
16 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
17 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
18 * ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
19 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
20 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
21 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
22 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
23 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
24 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
25 * SUCH DAMAGE.
26 */
27
28
29 #include "mpdecimal.h"
30
31 #include <assert.h>
32
33 #include "bits.h"
34 #include "constants.h"
35 #include "difradix2.h"
36 #include "numbertheory.h"
37 #include "umodarith.h"
38
39
40 /* Bignum: The actual transform routine (decimation in frequency). */
41
42
43 /*
44 * Generate index pairs (x, bitreverse(x)) and carry out the permutation.
45 * n must be a power of two.
46 * Algorithm due to Brent/Lehmann, see Joerg Arndt, "Matters Computational",
47 * Chapter 1.14.4. [http://www.jjj.de/fxt/]
48 */
49 static inline void
bitreverse_permute(mpd_uint_t a[],mpd_size_t n)50 bitreverse_permute(mpd_uint_t a[], mpd_size_t n)
51 {
52 mpd_size_t x = 0;
53 mpd_size_t r = 0;
54 mpd_uint_t t;
55
56 do { /* Invariant: r = bitreverse(x) */
57 if (r > x) {
58 t = a[x];
59 a[x] = a[r];
60 a[r] = t;
61 }
62 /* Flip trailing consecutive 1 bits and the first zero bit
63 * that absorbs a possible carry. */
64 x += 1;
65 /* Mirror the operation on r: Flip n_trailing_zeros(x)+1
66 high bits of r. */
67 r ^= (n - (n >> (mpd_bsf(x)+1)));
68 /* The loop invariant is preserved. */
69 } while (x < n);
70 }
71
72
73 /* Fast Number Theoretic Transform, decimation in frequency. */
74 void
fnt_dif2(mpd_uint_t a[],mpd_size_t n,struct fnt_params * tparams)75 fnt_dif2(mpd_uint_t a[], mpd_size_t n, struct fnt_params *tparams)
76 {
77 mpd_uint_t *wtable = tparams->wtable;
78 mpd_uint_t umod;
79 #ifdef PPRO
80 double dmod;
81 uint32_t dinvmod[3];
82 #endif
83 mpd_uint_t u0, u1, v0, v1;
84 mpd_uint_t w, w0, w1, wstep;
85 mpd_size_t m, mhalf;
86 mpd_size_t j, r;
87
88
89 assert(ispower2(n));
90 assert(n >= 4);
91
92 SETMODULUS(tparams->modnum);
93
94 /* m == n */
95 mhalf = n / 2;
96 for (j = 0; j < mhalf; j += 2) {
97
98 w0 = wtable[j];
99 w1 = wtable[j+1];
100
101 u0 = a[j];
102 v0 = a[j+mhalf];
103
104 u1 = a[j+1];
105 v1 = a[j+1+mhalf];
106
107 a[j] = addmod(u0, v0, umod);
108 v0 = submod(u0, v0, umod);
109
110 a[j+1] = addmod(u1, v1, umod);
111 v1 = submod(u1, v1, umod);
112
113 MULMOD2(&v0, w0, &v1, w1);
114
115 a[j+mhalf] = v0;
116 a[j+1+mhalf] = v1;
117
118 }
119
120 wstep = 2;
121 for (m = n/2; m >= 2; m>>=1, wstep<<=1) {
122
123 mhalf = m / 2;
124
125 /* j == 0 */
126 for (r = 0; r < n; r += 2*m) {
127
128 u0 = a[r];
129 v0 = a[r+mhalf];
130
131 u1 = a[m+r];
132 v1 = a[m+r+mhalf];
133
134 a[r] = addmod(u0, v0, umod);
135 v0 = submod(u0, v0, umod);
136
137 a[m+r] = addmod(u1, v1, umod);
138 v1 = submod(u1, v1, umod);
139
140 a[r+mhalf] = v0;
141 a[m+r+mhalf] = v1;
142 }
143
144 for (j = 1; j < mhalf; j++) {
145
146 w = wtable[j*wstep];
147
148 for (r = 0; r < n; r += 2*m) {
149
150 u0 = a[r+j];
151 v0 = a[r+j+mhalf];
152
153 u1 = a[m+r+j];
154 v1 = a[m+r+j+mhalf];
155
156 a[r+j] = addmod(u0, v0, umod);
157 v0 = submod(u0, v0, umod);
158
159 a[m+r+j] = addmod(u1, v1, umod);
160 v1 = submod(u1, v1, umod);
161
162 MULMOD2C(&v0, &v1, w);
163
164 a[r+j+mhalf] = v0;
165 a[m+r+j+mhalf] = v1;
166 }
167
168 }
169
170 }
171
172 bitreverse_permute(a, n);
173 }
174