1 | /* |
2 | * IBM Accurate Mathematical Library |
3 | * written by International Business Machines Corp. |
4 | * Copyright (C) 2001-2021 Free Software Foundation, Inc. |
5 | * |
6 | * This program is free software; you can redistribute it and/or modify |
7 | * it under the terms of the GNU Lesser General Public License as published by |
8 | * the Free Software Foundation; either version 2.1 of the License, or |
9 | * (at your option) any later version. |
10 | * |
11 | * This program is distributed in the hope that it will be useful, |
12 | * but WITHOUT ANY WARRANTY; without even the implied warranty of |
13 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
14 | * GNU Lesser General Public License for more details. |
15 | * |
16 | * You should have received a copy of the GNU Lesser General Public License |
17 | * along with this program; if not, see <https://www.gnu.org/licenses/>. |
18 | */ |
19 | /****************************************************************/ |
20 | /* MODULE_NAME: sincos32.c */ |
21 | /* */ |
22 | /* FUNCTIONS: ss32 */ |
23 | /* cc32 */ |
24 | /* c32 */ |
25 | /* sin32 */ |
26 | /* cos32 */ |
27 | /* mpsin */ |
28 | /* mpcos */ |
29 | /* mpranred */ |
30 | /* mpsin1 */ |
31 | /* mpcos1 */ |
32 | /* */ |
33 | /* FILES NEEDED: endian.h mpa.h sincos32.h */ |
34 | /* mpa.c */ |
35 | /* */ |
36 | /* Multi Precision sin() and cos() function with p=32 for sin()*/ |
37 | /* cos() arcsin() and arccos() routines */ |
38 | /* In addition mpranred() routine performs range reduction of */ |
39 | /* a double number x into multi precision number y, */ |
40 | /* such that y=x-n*pi/2, abs(y)<pi/4, n=0,+-1,+-2,.... */ |
41 | /****************************************************************/ |
42 | #include "endian.h" |
43 | #include "mpa.h" |
44 | #include "sincos32.h" |
45 | #include <math.h> |
46 | #include <math_private.h> |
47 | #include <stap-probe.h> |
48 | |
49 | #ifndef SECTION |
50 | # define SECTION |
51 | #endif |
52 | |
53 | /* Compute Multi-Precision sin() function for given p. Receive Multi Precision |
54 | number x and result stored at y. */ |
55 | static void |
56 | SECTION |
57 | ss32 (mp_no *x, mp_no *y, int p) |
58 | { |
59 | int i; |
60 | double a; |
61 | mp_no mpt1, x2, gor, sum, mpk = {1, {1.0}}; |
62 | for (i = 1; i <= p; i++) |
63 | mpk.d[i] = 0; |
64 | |
65 | __sqr (x, &x2, p); |
66 | __cpy (&oofac27, &gor, p); |
67 | __cpy (&gor, &sum, p); |
68 | for (a = 27.0; a > 1.0; a -= 2.0) |
69 | { |
70 | mpk.d[1] = a * (a - 1.0); |
71 | __mul (&gor, &mpk, &mpt1, p); |
72 | __cpy (&mpt1, &gor, p); |
73 | __mul (&x2, &sum, &mpt1, p); |
74 | __sub (&gor, &mpt1, &sum, p); |
75 | } |
76 | __mul (x, &sum, y, p); |
77 | } |
78 | |
79 | /* Compute Multi-Precision cos() function for given p. Receive Multi Precision |
80 | number x and result stored at y. */ |
81 | static void |
82 | SECTION |
83 | cc32 (mp_no *x, mp_no *y, int p) |
84 | { |
85 | int i; |
86 | double a; |
87 | mp_no mpt1, x2, gor, sum, mpk = {1, {1.0}}; |
88 | for (i = 1; i <= p; i++) |
89 | mpk.d[i] = 0; |
90 | |
91 | __sqr (x, &x2, p); |
92 | mpk.d[1] = 27.0; |
93 | __mul (&oofac27, &mpk, &gor, p); |
94 | __cpy (&gor, &sum, p); |
95 | for (a = 26.0; a > 2.0; a -= 2.0) |
96 | { |
97 | mpk.d[1] = a * (a - 1.0); |
98 | __mul (&gor, &mpk, &mpt1, p); |
99 | __cpy (&mpt1, &gor, p); |
100 | __mul (&x2, &sum, &mpt1, p); |
101 | __sub (&gor, &mpt1, &sum, p); |
102 | } |
103 | __mul (&x2, &sum, y, p); |
104 | } |
105 | |
106 | /* Compute both sin(x), cos(x) as Multi precision numbers. */ |
107 | void |
108 | SECTION |
109 | __c32 (mp_no *x, mp_no *y, mp_no *z, int p) |
110 | { |
111 | mp_no u, t, t1, t2, c, s; |
112 | int i; |
113 | __cpy (x, &u, p); |
114 | u.e = u.e - 1; |
115 | cc32 (&u, &c, p); |
116 | ss32 (&u, &s, p); |
117 | for (i = 0; i < 24; i++) |
118 | { |
119 | __mul (&c, &s, &t, p); |
120 | __sub (&s, &t, &t1, p); |
121 | __add (&t1, &t1, &s, p); |
122 | __sub (&__mptwo, &c, &t1, p); |
123 | __mul (&t1, &c, &t2, p); |
124 | __add (&t2, &t2, &c, p); |
125 | } |
126 | __sub (&__mpone, &c, y, p); |
127 | __cpy (&s, z, p); |
128 | } |
129 | |
130 | /* Compute sin() of double-length number (X + DX) as Multi Precision number and |
131 | return result as double. If REDUCE_RANGE is true, X is assumed to be the |
132 | original input and DX is ignored. */ |
133 | double |
134 | SECTION |
135 | __mpsin (double x, double dx, bool reduce_range) |
136 | { |
137 | double y; |
138 | mp_no a, b, c, s; |
139 | int n; |
140 | int p = 32; |
141 | |
142 | if (reduce_range) |
143 | { |
144 | n = __mpranred (x, &a, p); /* n is 0, 1, 2 or 3. */ |
145 | __c32 (&a, &c, &s, p); |
146 | } |
147 | else |
148 | { |
149 | n = -1; |
150 | __dbl_mp (x, &b, p); |
151 | __dbl_mp (dx, &c, p); |
152 | __add (&b, &c, &a, p); |
153 | if (x > 0.8) |
154 | { |
155 | __sub (&hp, &a, &b, p); |
156 | __c32 (&b, &s, &c, p); |
157 | } |
158 | else |
159 | __c32 (&a, &c, &s, p); /* b = sin(x+dx) */ |
160 | } |
161 | |
162 | /* Convert result based on which quarter of unit circle y is in. */ |
163 | switch (n) |
164 | { |
165 | case 1: |
166 | __mp_dbl (&c, &y, p); |
167 | break; |
168 | |
169 | case 3: |
170 | __mp_dbl (&c, &y, p); |
171 | y = -y; |
172 | break; |
173 | |
174 | case 2: |
175 | __mp_dbl (&s, &y, p); |
176 | y = -y; |
177 | break; |
178 | |
179 | /* Quadrant not set, so the result must be sin (X + DX), which is also in |
180 | S. */ |
181 | case 0: |
182 | default: |
183 | __mp_dbl (&s, &y, p); |
184 | } |
185 | LIBC_PROBE (slowsin, 3, &x, &dx, &y); |
186 | return y; |
187 | } |
188 | |
189 | /* Compute cos() of double-length number (X + DX) as Multi Precision number and |
190 | return result as double. If REDUCE_RANGE is true, X is assumed to be the |
191 | original input and DX is ignored. */ |
192 | double |
193 | SECTION |
194 | __mpcos (double x, double dx, bool reduce_range) |
195 | { |
196 | double y; |
197 | mp_no a, b, c, s; |
198 | int n; |
199 | int p = 32; |
200 | |
201 | if (reduce_range) |
202 | { |
203 | n = __mpranred (x, &a, p); /* n is 0, 1, 2 or 3. */ |
204 | __c32 (&a, &c, &s, p); |
205 | } |
206 | else |
207 | { |
208 | n = -1; |
209 | __dbl_mp (x, &b, p); |
210 | __dbl_mp (dx, &c, p); |
211 | __add (&b, &c, &a, p); |
212 | if (x > 0.8) |
213 | { |
214 | __sub (&hp, &a, &b, p); |
215 | __c32 (&b, &s, &c, p); |
216 | } |
217 | else |
218 | __c32 (&a, &c, &s, p); /* a = cos(x+dx) */ |
219 | } |
220 | |
221 | /* Convert result based on which quarter of unit circle y is in. */ |
222 | switch (n) |
223 | { |
224 | case 1: |
225 | __mp_dbl (&s, &y, p); |
226 | y = -y; |
227 | break; |
228 | |
229 | case 3: |
230 | __mp_dbl (&s, &y, p); |
231 | break; |
232 | |
233 | case 2: |
234 | __mp_dbl (&c, &y, p); |
235 | y = -y; |
236 | break; |
237 | |
238 | /* Quadrant not set, so the result must be cos (X + DX), which is also |
239 | stored in C. */ |
240 | case 0: |
241 | default: |
242 | __mp_dbl (&c, &y, p); |
243 | } |
244 | LIBC_PROBE (slowcos, 3, &x, &dx, &y); |
245 | return y; |
246 | } |
247 | |
248 | /* Perform range reduction of a double number x into multi precision number y, |
249 | such that y = x - n * pi / 2, abs (y) < pi / 4, n = 0, +-1, +-2, ... |
250 | Return int which indicates in which quarter of circle x is. */ |
251 | int |
252 | SECTION |
253 | __mpranred (double x, mp_no *y, int p) |
254 | { |
255 | number v; |
256 | double t, xn; |
257 | int i, k, n; |
258 | mp_no a, b, c; |
259 | |
260 | if (fabs (x) < 2.8e14) |
261 | { |
262 | t = (x * hpinv.d + toint.d); |
263 | xn = t - toint.d; |
264 | v.d = t; |
265 | n = v.i[LOW_HALF] & 3; |
266 | __dbl_mp (xn, &a, p); |
267 | __mul (&a, &hp, &b, p); |
268 | __dbl_mp (x, &c, p); |
269 | __sub (&c, &b, y, p); |
270 | return n; |
271 | } |
272 | else |
273 | { |
274 | /* If x is very big more precision required. */ |
275 | __dbl_mp (x, &a, p); |
276 | a.d[0] = 1.0; |
277 | k = a.e - 5; |
278 | if (k < 0) |
279 | k = 0; |
280 | b.e = -k; |
281 | b.d[0] = 1.0; |
282 | for (i = 0; i < p; i++) |
283 | b.d[i + 1] = toverp[i + k]; |
284 | __mul (&a, &b, &c, p); |
285 | t = c.d[c.e]; |
286 | for (i = 1; i <= p - c.e; i++) |
287 | c.d[i] = c.d[i + c.e]; |
288 | for (i = p + 1 - c.e; i <= p; i++) |
289 | c.d[i] = 0; |
290 | c.e = 0; |
291 | if (c.d[1] >= HALFRAD) |
292 | { |
293 | t += 1.0; |
294 | __sub (&c, &__mpone, &b, p); |
295 | __mul (&b, &hp, y, p); |
296 | } |
297 | else |
298 | __mul (&c, &hp, y, p); |
299 | n = (int) t; |
300 | if (x < 0) |
301 | { |
302 | y->d[0] = -y->d[0]; |
303 | n = -n; |
304 | } |
305 | return (n & 3); |
306 | } |
307 | } |
308 | |