ae747c9b |
1 | /* -*-c-*- |
2 | * |
ae747c9b |
3 | * Sqaring binary polynomials |
4 | * |
5 | * (c) 2000 Straylight/Edgeware |
6 | */ |
7 | |
45c0fd36 |
8 | /*----- Licensing notice --------------------------------------------------* |
ae747c9b |
9 | * |
10 | * This file is part of Catacomb. |
11 | * |
12 | * Catacomb is free software; you can redistribute it and/or modify |
13 | * it under the terms of the GNU Library General Public License as |
14 | * published by the Free Software Foundation; either version 2 of the |
15 | * License, or (at your option) any later version. |
45c0fd36 |
16 | * |
ae747c9b |
17 | * Catacomb is distributed in the hope that it will be useful, |
18 | * but WITHOUT ANY WARRANTY; without even the implied warranty of |
19 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
20 | * GNU Library General Public License for more details. |
45c0fd36 |
21 | * |
ae747c9b |
22 | * You should have received a copy of the GNU Library General Public |
23 | * License along with Catacomb; if not, write to the Free |
24 | * Software Foundation, Inc., 59 Temple Place - Suite 330, Boston, |
25 | * MA 02111-1307, USA. |
26 | */ |
27 | |
ae747c9b |
28 | /*----- Header files ------------------------------------------------------*/ |
29 | |
30 | #include "mpx.h" |
ceb3f0c0 |
31 | #include "gfx.h" |
ae747c9b |
32 | #include "gfx-sqr-tab.h" |
33 | |
34 | /*----- Static variables --------------------------------------------------*/ |
35 | |
4e66da02 |
36 | static const uint16 tab[256] = GFX_SQRTAB; |
ae747c9b |
37 | |
38 | /*----- Main code ---------------------------------------------------------*/ |
39 | |
40 | /* --- @gfx_sqr@ --- * |
41 | * |
42 | * Arguments: @mpw *dv, *dvl@ = destination vector base and limit |
43 | * @const mpw *av, *avl@ = argument vector base and limit |
44 | * |
45 | * Returns: --- |
46 | * |
47 | * Use: Performs squaring of binary polynomials. |
48 | */ |
49 | |
50 | void gfx_sqr(mpw *dv, mpw *dvl, const mpw *av, const mpw *avl) |
51 | { |
52 | mpd a = 0, aa = 0; |
53 | unsigned b = 0, bb = 0; |
54 | |
55 | /* --- Simple stuff --- */ |
56 | |
57 | if (dv >= dvl) |
58 | return; |
59 | MPX_SHRINK(av, avl); |
60 | |
61 | /* --- The main algorithm --- * |
62 | * |
63 | * Our method depends on the fact that, in a field of characteristic 2, we |
64 | * have that %$(a + b)^2 = a^2 + b^2$%. Thus, to square a polynomial, it's |
65 | * sufficient just to put a zero bit between each of the bits of the |
66 | * original argument. We use a precomputed table for this, and work on |
67 | * entire octets at a time. Life is more complicated because we've got to |
68 | * be careful of bizarre architectures which don't have words with a |
69 | * multiple of 8 bits in them. |
70 | */ |
71 | |
72 | for (;;) { |
73 | |
74 | /* --- Input buffering --- */ |
75 | |
76 | if (b < 8) { |
77 | if (av >= avl) |
78 | break; |
79 | a |= *av++ << b; |
80 | b += MPW_BITS; |
81 | } |
82 | |
83 | /* --- Do the work in the middle --- */ |
84 | |
85 | aa |= (mpd)(tab[U8(a)]) << bb; |
86 | bb += 16; |
87 | a >>= 8; |
88 | b -= 8; |
89 | |
90 | /* --- Output buffering --- */ |
91 | |
ceb3f0c0 |
92 | if (bb >= MPW_BITS) { |
ae747c9b |
93 | *dv++ = MPW(aa); |
94 | if (dv >= dvl) |
95 | return; |
96 | aa >>= MPW_BITS; |
97 | bb -= MPW_BITS; |
98 | } |
99 | } |
100 | |
101 | /* --- Flush the input buffer --- */ |
102 | |
103 | if (b) for (;;) { |
104 | aa |= (mpd)(tab[U8(a)]) << bb; |
105 | bb += 16; |
106 | if (bb > MPW_BITS) { |
107 | *dv++ = MPW(aa); |
108 | if (dv >= dvl) |
109 | return; |
110 | aa >>= MPW_BITS; |
111 | bb -= MPW_BITS; |
45c0fd36 |
112 | } |
ae747c9b |
113 | a >>= 8; |
114 | if (b <= 8) |
115 | break; |
116 | else |
117 | b -= 8; |
118 | } |
119 | |
120 | /* --- Flush the output buffer --- */ |
121 | |
122 | if (bb) for (;;) { |
123 | *dv++ = MPW(aa); |
124 | if (dv >= dvl) |
125 | return; |
126 | aa >>= MPW_BITS; |
127 | if (bb <= MPW_BITS) |
128 | break; |
129 | else |
130 | bb -= MPW_BITS; |
131 | } |
132 | |
133 | /* --- Zero the rest of everything --- */ |
134 | |
135 | MPX_ZERO(dv, dvl); |
136 | } |
137 | |
138 | /*----- Test rig ----------------------------------------------------------*/ |
139 | |
140 | #ifdef TEST_RIG |
141 | |
142 | #include <mLib/alloc.h> |
143 | #include <mLib/dstr.h> |
144 | #include <mLib/quis.h> |
145 | #include <mLib/testrig.h> |
146 | |
147 | #define ALLOC(v, vl, sz) do { \ |
148 | size_t _sz = (sz); \ |
149 | mpw *_vv = xmalloc(MPWS(_sz)); \ |
150 | mpw *_vvl = _vv + _sz; \ |
151 | (v) = _vv; \ |
152 | (vl) = _vvl; \ |
153 | } while (0) |
154 | |
155 | #define LOAD(v, vl, d) do { \ |
156 | const dstr *_d = (d); \ |
157 | mpw *_v, *_vl; \ |
158 | ALLOC(_v, _vl, MPW_RQ(_d->len)); \ |
159 | mpx_loadb(_v, _vl, _d->buf, _d->len); \ |
160 | (v) = _v; \ |
161 | (vl) = _vl; \ |
162 | } while (0) |
163 | |
164 | #define MAX(x, y) ((x) > (y) ? (x) : (y)) |
45c0fd36 |
165 | |
ae747c9b |
166 | static void dumpmp(const char *msg, const mpw *v, const mpw *vl) |
167 | { |
168 | fputs(msg, stderr); |
169 | MPX_SHRINK(v, vl); |
170 | while (v < vl) |
171 | fprintf(stderr, " %08lx", (unsigned long)*--vl); |
172 | fputc('\n', stderr); |
173 | } |
174 | |
175 | static int vsqr(dstr *v) |
176 | { |
177 | mpw *a, *al; |
178 | mpw *b, *bl; |
179 | mpw *d, *dl; |
180 | int ok = 1; |
181 | |
182 | LOAD(a, al, &v[0]); |
183 | LOAD(b, bl, &v[1]); |
184 | ALLOC(d, dl, 2 * (al - a)); |
185 | |
186 | gfx_sqr(d, dl, a, al); |
187 | if (!mpx_ueq(d, dl, b, bl)) { |
188 | fprintf(stderr, "\n*** vsqr failed\n"); |
45c0fd36 |
189 | dumpmp(" a", a, al); |
ae747c9b |
190 | dumpmp("expected", b, bl); |
191 | dumpmp(" result", d, dl); |
192 | ok = 0; |
193 | } |
194 | |
12ed8a1f |
195 | xfree(a); xfree(b); xfree(d); |
ae747c9b |
196 | return (ok); |
197 | } |
198 | |
199 | static test_chunk defs[] = { |
200 | { "sqr", vsqr, { &type_hex, &type_hex, 0 } }, |
201 | { 0, 0, { 0 } } |
202 | }; |
203 | |
204 | int main(int argc, char *argv[]) |
205 | { |
0f00dc4c |
206 | test_run(argc, argv, defs, SRCDIR"/t/gfx"); |
ae747c9b |
207 | return (0); |
208 | } |
209 | |
210 | #endif |
211 | |
212 | /*----- That's all, folks -------------------------------------------------*/ |