summaryrefslogtreecommitdiff
path: root/src/math/pow.c
blob: 82f684bdb0e363c6846d0f7dde445cf4b5818618 (plain)
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
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
/* origin: FreeBSD /usr/src/lib/msun/src/e_pow.c */
/*
 * ====================================================
 * Copyright (C) 2004 by Sun Microsystems, Inc. All rights reserved.
 *
 * Permission to use, copy, modify, and distribute this
 * software is freely granted, provided that this notice
 * is preserved.
 * ====================================================
 */
/* pow(x,y) return x**y
 *
 *                    n
 * Method:  Let x =  2   * (1+f)
 *      1. Compute and return log2(x) in two pieces:
 *              log2(x) = w1 + w2,
 *         where w1 has 53-24 = 29 bit trailing zeros.
 *      2. Perform y*log2(x) = n+y' by simulating muti-precision
 *         arithmetic, where |y'|<=0.5.
 *      3. Return x**y = 2**n*exp(y'*log2)
 *
 * Special cases:
 *      1.  (anything) ** 0  is 1
 *      2.  1 ** (anything)  is 1
 *      3.  (anything except 1) ** NAN is NAN
 *      4.  NAN ** (anything except 0) is NAN
 *      5.  +-(|x| > 1) **  +INF is +INF
 *      6.  +-(|x| > 1) **  -INF is +0
 *      7.  +-(|x| < 1) **  +INF is +0
 *      8.  +-(|x| < 1) **  -INF is +INF
 *      9.  -1          ** +-INF is 1
 *      10. +0 ** (+anything except 0, NAN)               is +0
 *      11. -0 ** (+anything except 0, NAN, odd integer)  is +0
 *      12. +0 ** (-anything except 0, NAN)               is +INF, raise divbyzero
 *      13. -0 ** (-anything except 0, NAN, odd integer)  is +INF, raise divbyzero
 *      14. -0 ** (+odd integer) is -0
 *      15. -0 ** (-odd integer) is -INF, raise divbyzero
 *      16. +INF ** (+anything except 0,NAN) is +INF
 *      17. +INF ** (-anything except 0,NAN) is +0
 *      18. -INF ** (+odd integer) is -INF
 *      19. -INF ** (anything) = -0 ** (-anything), (anything except odd integer)
 *      20. (anything) ** 1 is (anything)
 *      21. (anything) ** -1 is 1/(anything)
 *      22. (-anything) ** (integer) is (-1)**(integer)*(+anything**integer)
 *      23. (-anything except 0 and inf) ** (non-integer) is NAN
 *
 * Accuracy:
 *      pow(x,y) returns x**y nearly rounded. In particular
 *                      pow(integer,integer)
 *      always returns the correct integer provided it is
 *      representable.
 *
 * Constants :
 * The hexadecimal values are the intended ones for the following
 * constants. The decimal values may be used, provided that the
 * compiler will convert from decimal to binary accurately enough
 * to produce the hexadecimal values shown.
 */

#include "libm.h"

static const double
bp[]   = {1.0, 1.5,},
dp_h[] = { 0.0, 5.84962487220764160156e-01,}, /* 0x3FE2B803, 0x40000000 */
dp_l[] = { 0.0, 1.35003920212974897128e-08,}, /* 0x3E4CFDEB, 0x43CFD006 */
two53  =  9007199254740992.0, /* 0x43400000, 0x00000000 */
huge   =  1.0e300,
tiny   =  1.0e-300,
/* poly coefs for (3/2)*(log(x)-2s-2/3*s**3 */
L1 =  5.99999999999994648725e-01, /* 0x3FE33333, 0x33333303 */
L2 =  4.28571428578550184252e-01, /* 0x3FDB6DB6, 0xDB6FABFF */
L3 =  3.33333329818377432918e-01, /* 0x3FD55555, 0x518F264D */
L4 =  2.72728123808534006489e-01, /* 0x3FD17460, 0xA91D4101 */
L5 =  2.30660745775561754067e-01, /* 0x3FCD864A, 0x93C9DB65 */
L6 =  2.06975017800338417784e-01, /* 0x3FCA7E28, 0x4A454EEF */
P1 =  1.66666666666666019037e-01, /* 0x3FC55555, 0x5555553E */
P2 = -2.77777777770155933842e-03, /* 0xBF66C16C, 0x16BEBD93 */
P3 =  6.61375632143793436117e-05, /* 0x3F11566A, 0xAF25DE2C */
P4 = -1.65339022054652515390e-06, /* 0xBEBBBD41, 0xC5D26BF1 */
P5 =  4.13813679705723846039e-08, /* 0x3E663769, 0x72BEA4D0 */
lg2     =  6.93147180559945286227e-01, /* 0x3FE62E42, 0xFEFA39EF */
lg2_h   =  6.93147182464599609375e-01, /* 0x3FE62E43, 0x00000000 */
lg2_l   = -1.90465429995776804525e-09, /* 0xBE205C61, 0x0CA86C39 */
ovt     =  8.0085662595372944372e-017, /* -(1024-log2(ovfl+.5ulp)) */
cp      =  9.61796693925975554329e-01, /* 0x3FEEC709, 0xDC3A03FD =2/(3ln2) */
cp_h    =  9.61796700954437255859e-01, /* 0x3FEEC709, 0xE0000000 =(float)cp */
cp_l    = -7.02846165095275826516e-09, /* 0xBE3E2FE0, 0x145B01F5 =tail of cp_h*/
ivln2   =  1.44269504088896338700e+00, /* 0x3FF71547, 0x652B82FE =1/ln2 */
ivln2_h =  1.44269502162933349609e+00, /* 0x3FF71547, 0x60000000 =24b 1/ln2*/
ivln2_l =  1.92596299112661746887e-08; /* 0x3E54AE0B, 0xF85DDF44 =1/ln2 tail*/

double pow(double x, double y)
{
	double z,ax,z_h,z_l,p_h,p_l;
	double y1,t1,t2,r,s,t,u,v,w;
	int32_t i,j,k,yisint,n;
	int32_t hx,hy,ix,iy;
	uint32_t lx,ly;

	EXTRACT_WORDS(hx, lx, x);
	EXTRACT_WORDS(hy, ly, y);
	ix = hx & 0x7fffffff;
	iy = hy & 0x7fffffff;

	/* x**0 = 1, even if x is NaN */
	if ((iy|ly) == 0)
		return 1.0;
	/* 1**y = 1, even if y is NaN */
	if (hx == 0x3ff00000 && lx == 0)
		return 1.0;
	/* NaN if either arg is NaN */
	if (ix > 0x7ff00000 || (ix == 0x7ff00000 && lx != 0) ||
	    iy > 0x7ff00000 || (iy == 0x7ff00000 && ly != 0))
		return x + y;

	/* determine if y is an odd int when x < 0
	 * yisint = 0       ... y is not an integer
	 * yisint = 1       ... y is an odd int
	 * yisint = 2       ... y is an even int
	 */
	yisint = 0;
	if (hx < 0) {
		if (iy >= 0x43400000)
			yisint = 2; /* even integer y */
		else if (iy >= 0x3ff00000) {
			k = (iy>>20) - 0x3ff;  /* exponent */
			if (k > 20) {
				j = ly>>(52-k);
				if ((j<<(52-k)) == ly)
					yisint = 2 - (j&1);
			} else if (ly == 0) {
				j = iy>>(20-k);
				if ((j<<(20-k)) == iy)
					yisint = 2 - (j&1);
			}
		}
	}

	/* special value of y */
	if (ly == 0) {
		if (iy == 0x7ff00000) {  /* y is +-inf */
			if (((ix-0x3ff00000)|lx) == 0)  /* (-1)**+-inf is 1 */
				return 1.0;
			else if (ix >= 0x3ff00000) /* (|x|>1)**+-inf = inf,0 */
				return hy >= 0 ? y : 0.0;
			else if ((ix|lx) != 0)     /* (|x|<1)**+-inf = 0,inf if x!=0 */
				return hy >= 0 ? 0.0 : -y;
		}
		if (iy == 0x3ff00000) {    /* y is +-1 */
			if (hy >= 0)
				return x;
			y = 1/x;
#if FLT_EVAL_METHOD!=0
			{
				union {double f; uint64_t i;} u = {y};
				uint64_t i = u.i & -1ULL/2;
				if (i>>52 == 0 && (i&(i-1)))
					FORCE_EVAL((float)y);
			}
#endif
			return y;
		}
		if (hy == 0x40000000)    /* y is 2 */
			return x*x;
		if (hy == 0x3fe00000) {  /* y is 0.5 */
			if (hx >= 0)     /* x >= +0 */
				return sqrt(x);
		}
	}

	ax = fabs(x);
	/* special value of x */
	if (lx == 0) {
		if (ix == 0x7ff00000 || ix == 0 || ix == 0x3ff00000) { /* x is +-0,+-inf,+-1 */
			z = ax;
			if (hy < 0)   /* z = (1/|x|) */
				z = 1.0/z;
			if (hx < 0) {
				if (((ix-0x3ff00000)|yisint) == 0) {
					z = (z-z)/(z-z); /* (-1)**non-int is NaN */
				} else if (yisint == 1)
					z = -z;          /* (x<0)**odd = -(|x|**odd) */
			}
			return z;
		}
	}

	s = 1.0; /* sign of result */
	if (hx < 0) {
		if (yisint == 0) /* (x<0)**(non-int) is NaN */
			return (x-x)/(x-x);
		if (yisint == 1) /* (x<0)**(odd int) */
			s = -1.0;
	}

	/* |y| is huge */
	if (iy > 0x41e00000) { /* if |y| > 2**31 */
		if (iy > 0x43f00000) {  /* if |y| > 2**64, must o/uflow */
			if (ix <= 0x3fefffff)
				return hy < 0 ? huge*huge : tiny*tiny;
			if (ix >= 0x3ff00000)
				return hy > 0 ? huge*huge : tiny*tiny;
		}
		/* over/underflow if x is not close to one */
		if (ix < 0x3fefffff)
			return hy < 0 ? s*huge*huge : s*tiny*tiny;
		if (ix > 0x3ff00000)
			return hy > 0 ? s*huge*huge : s*tiny*tiny;
		/* now |1-x| is tiny <= 2**-20, suffice to compute
		   log(x) by x-x^2/2+x^3/3-x^4/4 */
		t = ax - 1.0;       /* t has 20 trailing zeros */
		w = (t*t)*(0.5 - t*(0.3333333333333333333333-t*0.25));
		u = ivln2_h*t;      /* ivln2_h has 21 sig. bits */
		v = t*ivln2_l - w*ivln2;
		t1 = u + v;
		SET_LOW_WORD(t1, 0);
		t2 = v - (t1-u);
	} else {
		double ss,s2,s_h,s_l,t_h,t_l;
		n = 0;
		/* take care subnormal number */
		if (ix < 0x00100000) {
			ax *= two53;
			n -= 53;
			GET_HIGH_WORD(ix,ax);
		}
		n += ((ix)>>20) - 0x3ff;
		j = ix & 0x000fffff;
		/* determine interval */
		ix = j | 0x3ff00000;   /* normalize ix */
		if (j <= 0x3988E)      /* |x|<sqrt(3/2) */
			k = 0;
		else if (j < 0xBB67A)  /* |x|<sqrt(3)   */
			k = 1;
		else {
			k = 0;
			n += 1;
			ix -= 0x00100000;
		}
		SET_HIGH_WORD(ax, ix);

		/* compute ss = s_h+s_l = (x-1)/(x+1) or (x-1.5)/(x+1.5) */
		u = ax - bp[k];        /* bp[0]=1.0, bp[1]=1.5 */
		v = 1.0/(ax+bp[k]);
		ss = u*v;
		s_h = ss;
		SET_LOW_WORD(s_h, 0);
		/* t_h=ax+bp[k] High */
		t_h = 0.0;
		SET_HIGH_WORD(t_h, ((ix>>1)|0x20000000) + 0x00080000 + (k<<18));
		t_l = ax - (t_h-bp[k]);
		s_l = v*((u-s_h*t_h)-s_h*t_l);
		/* compute log(ax) */
		s2 = ss*ss;
		r = s2*s2*(L1+s2*(L2+s2*(L3+s2*(L4+s2*(L5+s2*L6)))));
		r += s_l*(s_h+ss);
		s2 = s_h*s_h;
		t_h = 3.0 + s2 + r;
		SET_LOW_WORD(t_h, 0);
		t_l = r - ((t_h-3.0)-s2);
		/* u+v = ss*(1+...) */
		u = s_h*t_h;
		v = s_l*t_h + t_l*ss;
		/* 2/(3log2)*(ss+...) */
		p_h = u + v;
		SET_LOW_WORD(p_h, 0);
		p_l = v - (p_h-u);
		z_h = cp_h*p_h;        /* cp_h+cp_l = 2/(3*log2) */
		z_l = cp_l*p_h+p_l*cp + dp_l[k];
		/* log2(ax) = (ss+..)*2/(3*log2) = n + dp_h + z_h + z_l */
		t = (double)n;
		t1 = ((z_h + z_l) + dp_h[k]) + t;
		SET_LOW_WORD(t1, 0);
		t2 = z_l - (((t1 - t) - dp_h[k]) - z_h);
	}

	/* split up y into y1+y2 and compute (y1+y2)*(t1+t2) */
	y1 = y;
	SET_LOW_WORD(y1, 0);
	p_l = (y-y1)*t1 + y*t2;
	p_h = y1*t1;
	z = p_l + p_h;
	EXTRACT_WORDS(j, i, z);
	if (j >= 0x40900000) {                      /* z >= 1024 */
		if (((j-0x40900000)|i) != 0)        /* if z > 1024 */
			return s*huge*huge;         /* overflow */
		if (p_l + ovt > z - p_h)
			return s*huge*huge;         /* overflow */
	} else if ((j&0x7fffffff) >= 0x4090cc00) {  /* z <= -1075 */  // FIXME: instead of abs(j) use unsigned j
		if (((j-0xc090cc00)|i) != 0)        /* z < -1075 */
			return s*tiny*tiny;         /* underflow */
		if (p_l <= z - p_h)
			return s*tiny*tiny;         /* underflow */
	}
	/*
	 * compute 2**(p_h+p_l)
	 */
	i = j & 0x7fffffff;
	k = (i>>20) - 0x3ff;
	n = 0;
	if (i > 0x3fe00000) {  /* if |z| > 0.5, set n = [z+0.5] */
		n = j + (0x00100000>>(k+1));
		k = ((n&0x7fffffff)>>20) - 0x3ff;  /* new k for n */
		t = 0.0;
		SET_HIGH_WORD(t, n & ~(0x000fffff>>k));
		n = ((n&0x000fffff)|0x00100000)>>(20-k);
		if (j < 0)
			n = -n;
		p_h -= t;
	}
	t = p_l + p_h;
	SET_LOW_WORD(t, 0);
	u = t*lg2_h;
	v = (p_l-(t-p_h))*lg2 + t*lg2_l;
	z = u + v;
	w = v - (z-u);
	t = z*z;
	t1 = z - t*(P1+t*(P2+t*(P3+t*(P4+t*P5))));
	r = (z*t1)/(t1-2.0) - (w + z*w);
	z = 1.0 - (r-z);
	GET_HIGH_WORD(j, z);
	j += n<<20;
	if ((j>>20) <= 0)  /* subnormal output */
		z = scalbn(z,n);
	else
		SET_HIGH_WORD(z, j);
	return s*z;
}