1 /****************************************************************
3 * The author of this software is David M. Gay.
5 * Copyright (c) 1991, 2000, 2001 by Lucent Technologies.
7 * Permission to use, copy, modify, and distribute this software for any
8 * purpose without fee is hereby granted, provided that this entire notice
9 * is included in all copies of any software which is or includes a copy
10 * or modification of this software and in all copies of the supporting
11 * documentation for such software.
13 * THIS SOFTWARE IS BEING PROVIDED "AS IS", WITHOUT ANY EXPRESS OR IMPLIED
14 * WARRANTY. IN PARTICULAR, NEITHER THE AUTHOR NOR LUCENT MAKES ANY
15 * REPRESENTATION OR WARRANTY OF ANY KIND CONCERNING THE MERCHANTABILITY
16 * OF THIS SOFTWARE OR ITS FITNESS FOR ANY PARTICULAR PURPOSE.
18 ***************************************************************/
21 #define freedtoa __freedtoa
24 #define Omit_Private_Memory
25 #define MULTIPLE_THREADS 1
26 /* Lock 0 is not used because of USE_MALLOC, Lock 1 protects a lazy-initialized table */
27 #define ACQUIRE_DTOA_LOCK(n)
28 #define FREE_DTOA_LOCK(n)
30 /* Please send bug reports to David M. Gay (dmg at acm dot org,
31 * with " at " changed at "@" and " dot " changed to "."). */
33 /* On a machine with IEEE extended-precision registers, it is
34 * necessary to specify double-precision (53-bit) rounding precision
35 * before invoking strtod or dtoa. If the machine uses (the equivalent
36 * of) Intel 80x87 arithmetic, the call
37 * _control87(PC_53, MCW_PC);
38 * does this with many compilers. Whether this or another call is
39 * appropriate depends on the compiler; for this to work, it may be
40 * necessary to #include "float.h" or another system-dependent header
44 /* strtod for IEEE-, VAX-, and IBM-arithmetic machines.
46 * This strtod returns a nearest machine number to the input decimal
47 * string (or sets errno to ERANGE). With IEEE arithmetic, ties are
48 * broken by the IEEE round-even rule. Otherwise ties are broken by
49 * biased rounding (add half and chop).
51 * Inspired loosely by William D. Clinger's paper "How to Read Floating
52 * Point Numbers Accurately" [Proc. ACM SIGPLAN '90, pp. 92-101].
56 * 1. We only require IEEE, IBM, or VAX double-precision
57 * arithmetic (not IEEE double-extended).
58 * 2. We get by with floating-point arithmetic in a case that
59 * Clinger missed -- when we're computing d * 10^n
60 * for a small integer d and the integer n is not too
61 * much larger than 22 (the maximum integer k for which
62 * we can represent 10^k exactly), we may be able to
63 * compute (d*10^k) * 10^(e-k) with just one roundoff.
64 * 3. Rather than a bit-at-a-time adjustment of the binary
65 * result in the hard case, we use floating-point
66 * arithmetic to determine the adjustment to within
67 * one bit; only in really hard cases do we need to
68 * compute a second residual.
69 * 4. Because of 3., we don't need a large table of powers of 10
70 * for ten-to-e (just some small tables, e.g. of 10^k
75 * #define IEEE_8087 for IEEE-arithmetic machines where the least
76 * significant byte has the lowest address.
77 * #define IEEE_MC68k for IEEE-arithmetic machines where the most
78 * significant byte has the lowest address.
79 * #define Long int on machines with 32-bit ints and 64-bit longs.
80 * #define IBM for IBM mainframe-style floating-point arithmetic.
81 * #define VAX for VAX-style floating-point arithmetic (D_floating).
82 * #define No_leftright to omit left-right logic in fast floating-point
83 * computation of dtoa.
84 * #define Honor_FLT_ROUNDS if FLT_ROUNDS can assume the values 2 or 3
85 * and strtod and dtoa should round accordingly.
86 * #define Check_FLT_ROUNDS if FLT_ROUNDS can assume the values 2 or 3
87 * and Honor_FLT_ROUNDS is not #defined.
88 * #define RND_PRODQUOT to use rnd_prod and rnd_quot (assembly routines
89 * that use extended-precision instructions to compute rounded
90 * products and quotients) with IBM.
91 * #define ROUND_BIASED for IEEE-format with biased rounding.
92 * #define Inaccurate_Divide for IEEE-format with correctly rounded
93 * products but inaccurate quotients, e.g., for Intel i860.
94 * #define NO_LONG_LONG on machines that do not have a "long long"
95 * integer type (of >= 64 bits). On such machines, you can
96 * #define Just_16 to store 16 bits per 32-bit Long when doing
97 * high-precision integer arithmetic. Whether this speeds things
98 * up or slows things down depends on the machine and the number
99 * being converted. If long long is available and the name is
100 * something other than "long long", #define Llong to be the name,
101 * and if "unsigned Llong" does not work as an unsigned version of
102 * Llong, #define #ULLong to be the corresponding unsigned type.
103 * #define KR_headers for old-style C function headers.
104 * #define Bad_float_h if your system lacks a float.h or if it does not
105 * define some or all of DBL_DIG, DBL_MAX_10_EXP, DBL_MAX_EXP,
106 * FLT_RADIX, FLT_ROUNDS, and DBL_MAX.
107 * #define MALLOC your_malloc, where your_malloc(n) acts like malloc(n)
108 * if memory is available and otherwise does something you deem
109 * appropriate. If MALLOC is undefined, malloc will be invoked
110 * directly -- and assumed always to succeed.
111 * #define Omit_Private_Memory to omit logic (added Jan. 1998) for making
112 * memory allocations from a private pool of memory when possible.
113 * When used, the private pool is PRIVATE_MEM bytes long: 2304 bytes,
114 * unless #defined to be a different length. This default length
115 * suffices to get rid of MALLOC calls except for unusual cases,
116 * such as decimal-to-binary conversion of a very long string of
117 * digits. The longest string dtoa can return is about 751 bytes
118 * long. For conversions by strtod of strings of 800 digits and
119 * all dtoa conversions in single-threaded executions with 8-byte
120 * pointers, PRIVATE_MEM >= 7400 appears to suffice; with 4-byte
121 * pointers, PRIVATE_MEM >= 7112 appears adequate.
122 * #define INFNAN_CHECK on IEEE systems to cause strtod to check for
123 * Infinity and NaN (case insensitively). On some systems (e.g.,
124 * some HP systems), it may be necessary to #define NAN_WORD0
125 * appropriately -- to the most significant word of a quiet NaN.
126 * (On HP Series 700/800 machines, -DNAN_WORD0=0x7ff40000 works.)
127 * When INFNAN_CHECK is #defined and No_Hex_NaN is not #defined,
128 * strtod also accepts (case insensitively) strings of the form
129 * NaN(x), where x is a string of hexadecimal digits and spaces;
130 * if there is only one string of hexadecimal digits, it is taken
131 * for the 52 fraction bits of the resulting NaN; if there are two
132 * or more strings of hex digits, the first is for the high 20 bits,
133 * the second and subsequent for the low 32 bits, with intervening
134 * white space ignored; but if this results in none of the 52
135 * fraction bits being on (an IEEE Infinity symbol), then NAN_WORD0
136 * and NAN_WORD1 are used instead.
137 * #define MULTIPLE_THREADS if the system offers preemptively scheduled
138 * multiple threads. In this case, you must provide (or suitably
139 * #define) two locks, acquired by ACQUIRE_DTOA_LOCK(n) and freed
140 * by FREE_DTOA_LOCK(n) for n = 0 or 1. (The second lock, accessed
141 * in pow5mult, ensures lazy evaluation of only one copy of high
142 * powers of 5; omitting this lock would introduce a small
143 * probability of wasting memory, but would otherwise be harmless.)
144 * You must also invoke freedtoa(s) to free the value s returned by
145 * dtoa. You may do so whether or not MULTIPLE_THREADS is #defined.
146 * #define NO_IEEE_Scale to disable new (Feb. 1997) logic in strtod that
147 * avoids underflows on inputs whose result does not underflow.
148 * If you #define NO_IEEE_Scale on a machine that uses IEEE-format
149 * floating-point numbers and flushes underflows to zero rather
150 * than implementing gradual underflow, then you must also #define
152 * #define YES_ALIAS to permit aliasing certain double values with
153 * arrays of ULongs. This leads to slightly better code with
154 * some compilers and was always used prior to 19990916, but it
155 * is not strictly legal and can cause trouble with aggressively
156 * optimizing compilers (e.g., gcc 2.95.1 under -O2).
157 * #define USE_LOCALE to use the current locale's decimal_point value.
158 * #define SET_INEXACT if IEEE arithmetic is being used and extra
159 * computation should be done to set the inexact flag when the
160 * result is inexact and avoid setting inexact when the result
161 * is exact. In this case, dtoa.c must be compiled in
162 * an environment, perhaps provided by #include "dtoa.c" in a
163 * suitable wrapper, that defines two functions,
164 * int get_inexact(void);
165 * void clear_inexact(void);
166 * such that get_inexact() returns a nonzero value if the
167 * inexact bit is already set, and clear_inexact() sets the
168 * inexact bit to 0. When SET_INEXACT is #defined, strtod
169 * also does extra computations to set the underflow and overflow
170 * flags when appropriate (i.e., when the result is tiny and
171 * inexact or when it is a numeric value rounded to +-infinity).
172 * #define NO_ERRNO if strtod should not assign errno = ERANGE when
173 * the result overflows to +-Infinity or underflows to 0.
175 #if defined(i386) || defined(mips) && defined(MIPSEL) || defined (__arm__)
179 #elif defined(__x86_64__) || defined(__alpha__)
183 #elif defined(__ia64)
191 #elif defined(__hppa)
200 #define ULong guint32
204 #define Bug(x) {fprintf(stderr, "%s\n", x); exit(1);}
217 extern char *MALLOC();
219 extern void *MALLOC(size_t);
222 #define MALLOC malloc
225 #define Omit_Private_Memory
226 #ifndef Omit_Private_Memory
228 #define PRIVATE_MEM 2304
230 #define PRIVATE_mem ((PRIVATE_MEM+sizeof(double)-1)/sizeof(double))
231 static double private_mem[PRIVATE_mem], *pmem_next = private_mem;
235 #undef Avoid_Underflow
249 #define DBL_MAX_10_EXP 308
250 #define DBL_MAX_EXP 1024
252 #endif /*IEEE_Arith*/
256 #define DBL_MAX_10_EXP 75
257 #define DBL_MAX_EXP 63
259 #define DBL_MAX 7.2370055773322621e+75
264 #define DBL_MAX_10_EXP 38
265 #define DBL_MAX_EXP 127
267 #define DBL_MAX 1.7014118346046923e+38
271 #define LONG_MAX 2147483647
274 #else /* ifndef Bad_float_h */
276 #endif /* Bad_float_h */
288 #define CONST /* blank */
294 #if defined(IEEE_8087) + defined(IEEE_MC68k) + defined(VAX) + defined(IBM) != 1
295 Exactly one of IEEE_8087, IEEE_MC68k, VAX, or IBM should be defined.
298 typedef union { double d; ULong L[2]; } U;
303 #define word0(x) ((ULong *)&x)[1]
304 #define word1(x) ((ULong *)&x)[0]
306 #define word0(x) ((ULong *)&x)[0]
307 #define word1(x) ((ULong *)&x)[1]
311 #define word0(x) ((U*)&x)->L[1]
312 #define word1(x) ((U*)&x)->L[0]
314 #define word0(x) ((U*)&x)->L[0]
315 #define word1(x) ((U*)&x)->L[1]
317 #define dval(x) ((U*)&x)->d
320 /* The following definition of Storeinc is appropriate for MIPS processors.
321 * An alternative that might be better on some machines is
322 * #define Storeinc(a,b,c) (*a++ = b << 16 | c & 0xffff)
324 #if defined(IEEE_8087) + defined(VAX)
325 #define Storeinc(a,b,c) do { (((unsigned short *)a)[1] = (unsigned short)b, \
326 ((unsigned short *)a)[0] = (unsigned short)c, a++) } while (0)
328 #define Storeinc(a,b,c) do { (((unsigned short *)a)[0] = (unsigned short)b, \
329 ((unsigned short *)a)[1] = (unsigned short)c, a++) } while (0)
332 /* #define P DBL_MANT_DIG */
333 /* Ten_pmax = floor(P*log(2)/log(5)) */
334 /* Bletch = (highest power of 2 < DBL_MAX_10_EXP) / 16 */
335 /* Quick_max = floor((P-1)*log(FLT_RADIX)/log(10) - 1) */
336 /* Int_max = floor(P*log(FLT_RADIX)/log(10) - 1) */
340 #define Exp_shift1 20
341 #define Exp_msk1 0x100000
342 #define Exp_msk11 0x100000
343 #define Exp_mask 0x7ff00000
347 #define Exp_1 0x3ff00000
348 #define Exp_11 0x3ff00000
350 #define Frac_mask 0xfffff
351 #define Frac_mask1 0xfffff
354 #define Bndry_mask 0xfffff
355 #define Bndry_mask1 0xfffff
357 #define Sign_bit 0x80000000
363 #ifndef NO_IEEE_Scale
364 #define Avoid_Underflow
365 #ifdef Flush_Denorm /* debugging option */
366 #undef Sudden_Underflow
372 #define Flt_Rounds FLT_ROUNDS
376 #endif /*Flt_Rounds*/
378 #ifdef Honor_FLT_ROUNDS
379 #define Rounding rounding
380 #undef Check_FLT_ROUNDS
381 #define Check_FLT_ROUNDS
383 #define Rounding Flt_Rounds
386 #else /* ifndef IEEE_Arith */
387 #undef Check_FLT_ROUNDS
388 #undef Honor_FLT_ROUNDS
390 #undef Sudden_Underflow
391 #define Sudden_Underflow
396 #define Exp_shift1 24
397 #define Exp_msk1 0x1000000
398 #define Exp_msk11 0x1000000
399 #define Exp_mask 0x7f000000
402 #define Exp_1 0x41000000
403 #define Exp_11 0x41000000
404 #define Ebits 8 /* exponent has 7 bits, but 8 is the right value in b2d */
405 #define Frac_mask 0xffffff
406 #define Frac_mask1 0xffffff
409 #define Bndry_mask 0xefffff
410 #define Bndry_mask1 0xffffff
412 #define Sign_bit 0x80000000
414 #define Tiny0 0x100000
423 #define Exp_msk1 0x80
424 #define Exp_msk11 0x800000
425 #define Exp_mask 0x7f80
428 #define Exp_1 0x40800000
429 #define Exp_11 0x4080
431 #define Frac_mask 0x7fffff
432 #define Frac_mask1 0xffff007f
435 #define Bndry_mask 0xffff007f
436 #define Bndry_mask1 0xffff007f
438 #define Sign_bit 0x8000
444 #endif /* IBM, VAX */
445 #endif /* IEEE_Arith */
452 #define rounded_product(a,b) a = rnd_prod(a, b)
453 #define rounded_quotient(a,b) a = rnd_quot(a, b)
455 extern double rnd_prod(), rnd_quot();
457 extern double rnd_prod(double, double), rnd_quot(double, double);
460 #define rounded_product(a,b) a *= b
461 #define rounded_quotient(a,b) a /= b
464 #define Big0 (Frac_mask1 | Exp_msk1*(DBL_MAX_EXP+Bias-1))
465 #define Big1 0xffffffff
472 #define FFFFFFFF ((((unsigned long)0xffff)<<16)|(unsigned long)0xffff)
474 #define FFFFFFFF 0xffffffffUL
481 /* When Pack_32 is not defined, we store 16 bits per 32-bit Long.
482 * This makes some inner loops simpler and sometimes saves work
483 * during multiplications, but it often seems to make things slightly
484 * slower. Hence the default is now to store 32 bits per Long.
487 #else /* long long available */
489 #define Llong long long
492 #define ULLong unsigned Llong
494 #endif /* NO_LONG_LONG */
496 #ifndef MULTIPLE_THREADS
497 #define ACQUIRE_DTOA_LOCK(n) /*nothing*/
498 #define FREE_DTOA_LOCK(n) /*nothing*/
504 extern "C" double strtod(const char *s00, char **se);
505 extern "C" char *dtoa(double d, int mode, int ndigits,
506 int *decpt, int *sign, char **rve);
512 int k, maxwds, sign, wds;
516 typedef struct Bigint Bigint;
518 static Bigint *freelist[Kmax+1];
530 #ifndef Omit_Private_Memory
534 ACQUIRE_DTOA_LOCK(0);
535 if ((rv = freelist[k])) {
536 freelist[k] = rv->next;
540 #ifdef Omit_Private_Memory
541 rv = (Bigint *)MALLOC(sizeof(Bigint) + (x-1)*sizeof(ULong));
543 len = (sizeof(Bigint) + (x-1)*sizeof(ULong) + sizeof(double) - 1)
545 if (pmem_next - private_mem + len <= PRIVATE_mem) {
546 rv = (Bigint*)pmem_next;
550 rv = (Bigint*)MALLOC(len*sizeof(double));
556 rv->sign = rv->wds = 0;
568 #ifdef Omit_Private_Memory
572 ACQUIRE_DTOA_LOCK(0);
573 v->next = freelist[v->k];
580 #define Bcopy(x,y) memcpy((char *)&x->sign, (char *)&y->sign, \
581 y->wds*sizeof(Long) + 2*sizeof(int))
586 (b, m, a) Bigint *b; int m, a;
588 (Bigint *b, int m, int a) /* multiply by m and add a */
609 y = *x * (ULLong)m + carry;
615 y = (xi & 0xffff) * m + carry;
616 z = (xi >> 16) * m + (y >> 16);
618 *x++ = (z << 16) + (y & 0xffff);
628 if (wds >= b->maxwds) {
643 (s, nd0, nd, y9) CONST char *s; int nd0, nd; ULong y9;
645 (CONST char *s, int nd0, int nd, ULong y9)
653 for(k = 0, y = 1; x > y; y <<= 1, k++) ;
660 b->x[0] = y9 & 0xffff;
661 b->wds = (b->x[1] = y9 >> 16) ? 2 : 1;
667 do b = multadd(b, 10, *s++ - '0');
674 b = multadd(b, 10, *s++ - '0');
681 (x) register ULong x;
688 if (!(x & 0xffff0000)) {
692 if (!(x & 0xff000000)) {
696 if (!(x & 0xf0000000)) {
700 if (!(x & 0xc0000000)) {
704 if (!(x & 0x80000000)) {
706 if (!(x & 0x40000000))
721 register ULong x = *y;
779 (a, b) Bigint *a, *b;
781 (Bigint *a, Bigint *b)
786 ULong *x, *xa, *xae, *xb, *xbe, *xc, *xc0;
797 if (a->wds < b->wds) {
809 for(x = c->x, xa = x + wc; x < xa; x++)
817 for(; xb < xbe; xc0++) {
823 z = *x++ * (ULLong)y + *xc + carry;
825 *xc++ = z & FFFFFFFF;
833 for(; xb < xbe; xb++, xc0++) {
834 if (y = *xb & 0xffff) {
839 z = (*x & 0xffff) * y + (*xc & 0xffff) + carry;
841 z2 = (*x++ >> 16) * y + (*xc >> 16) + carry;
854 z = (*x & 0xffff) * y + (*xc >> 16) + carry;
857 z2 = (*x++ >> 16) * y + (*xc & 0xffff) + carry;
865 for(; xb < xbe; xc0++) {
871 z = *x++ * y + *xc + carry;
881 for(xc0 = c->x, xc = xc0 + wc; wc > 0 && !*--xc; --wc) ;
891 (b, k) Bigint *b; int k;
896 Bigint *b1, *p5, *p51;
898 static int p05[3] = { 5, 25, 125 };
901 b = multadd(b, p05[i-1], 0);
907 #ifdef MULTIPLE_THREADS
908 ACQUIRE_DTOA_LOCK(1);
927 if (!(p51 = p5->next)) {
928 #ifdef MULTIPLE_THREADS
929 ACQUIRE_DTOA_LOCK(1);
930 if (!(p51 = p5->next)) {
931 p51 = p5->next = mult(p5,p5);
936 p51 = p5->next = mult(p5,p5);
948 (b, k) Bigint *b; int k;
955 ULong *x, *x1, *xe, z;
964 for(i = b->maxwds; n1 > i; i <<= 1)
968 for(i = 0; i < n; i++)
989 *x1++ = *x << k & 0xffff | z;
1008 (a, b) Bigint *a, *b;
1010 (Bigint *a, Bigint *b)
1013 ULong *xa, *xa0, *xb, *xb0;
1019 if (i > 1 && !a->x[i-1])
1020 Bug("cmp called with a->x[a->wds-1] == 0");
1021 if (j > 1 && !b->x[j-1])
1022 Bug("cmp called with b->x[b->wds-1] == 0");
1032 return *xa < *xb ? -1 : 1;
1042 (a, b) Bigint *a, *b;
1044 (Bigint *a, Bigint *b)
1049 ULong *xa, *xae, *xb, *xbe, *xc;
1086 y = (ULLong)*xa++ - *xb++ - borrow;
1087 borrow = y >> 32 & (ULong)1;
1088 *xc++ = y & FFFFFFFF;
1093 borrow = y >> 32 & (ULong)1;
1094 *xc++ = y & FFFFFFFF;
1099 y = (*xa & 0xffff) - (*xb & 0xffff) - borrow;
1100 borrow = (y & 0x10000) >> 16;
1101 z = (*xa++ >> 16) - (*xb++ >> 16) - borrow;
1102 borrow = (z & 0x10000) >> 16;
1107 y = (*xa & 0xffff) - borrow;
1108 borrow = (y & 0x10000) >> 16;
1109 z = (*xa++ >> 16) - borrow;
1110 borrow = (z & 0x10000) >> 16;
1115 y = *xa++ - *xb++ - borrow;
1116 borrow = (y & 0x10000) >> 16;
1122 borrow = (y & 0x10000) >> 16;
1144 L = (word0(x) & Exp_mask) - (P-1)*Exp_msk1;
1145 #ifndef Avoid_Underflow
1146 #ifndef Sudden_Underflow
1155 #ifndef Avoid_Underflow
1156 #ifndef Sudden_Underflow
1159 L = -L >> Exp_shift;
1160 if (L < Exp_shift) {
1161 word0(a) = 0x80000 >> L;
1167 word1(a) = L >= 31 ? 1 : 1 << 31 - L;
1178 (a, e) Bigint *a; int *e;
1183 ULong *xa, *xa0, w, y, z;
1197 if (!y) Bug("zero y in b2d");
1203 d0 = Exp_1 | (y >> (Ebits - k));
1204 w = xa > xa0 ? *--xa : 0;
1205 d1 = y << ((32-Ebits) + k) | (w >> (Ebits - k));
1208 z = xa > xa0 ? *--xa : 0;
1210 d0 = Exp_1 | y << k | (z >> (32 - k));
1211 y = xa > xa0 ? *--xa : 0;
1212 d1 = z << k | (y >> (32 - k));
1219 if (k < Ebits + 16) {
1220 z = xa > xa0 ? *--xa : 0;
1221 d0 = Exp_1 | y << k - Ebits | z >> Ebits + 16 - k;
1222 w = xa > xa0 ? *--xa : 0;
1223 y = xa > xa0 ? *--xa : 0;
1224 d1 = z << k + 16 - Ebits | w << k - Ebits | y >> 16 + Ebits - k;
1227 z = xa > xa0 ? *--xa : 0;
1228 w = xa > xa0 ? *--xa : 0;
1230 d0 = Exp_1 | y << k + 16 | z << k | w >> 16 - k;
1231 y = xa > xa0 ? *--xa : 0;
1232 d1 = w << k + 16 | y << k;
1236 word0(d) = d0 >> 16 | d0 << 16;
1237 word1(d) = d1 >> 16 | d1 << 16;
1248 (d, e, bits) double d; int *e, *bits;
1250 (double d, int *e, int *bits)
1256 #ifndef Sudden_Underflow
1261 d0 = word0(d) >> 16 | word0(d) << 16;
1262 d1 = word1(d) >> 16 | word1(d) << 16;
1276 d0 &= 0x7fffffff; /* clear sign bit, which we ignore */
1277 #ifdef Sudden_Underflow
1278 de = (int)(d0 >> Exp_shift);
1283 if ((de = (int)(d0 >> Exp_shift)))
1288 if ((k = lo0bits(&y))) {
1289 x[0] = y | (z << (32 - k));
1294 #ifndef Sudden_Underflow
1297 b->wds = (x[1] = z) ? 2 : 1;
1302 Bug("Zero passed to d2b");
1306 #ifndef Sudden_Underflow
1314 if (k = lo0bits(&y))
1316 x[0] = y | z << 32 - k & 0xffff;
1317 x[1] = z >> k - 16 & 0xffff;
1323 x[1] = y >> 16 | z << 16 - k & 0xffff;
1324 x[2] = z >> k & 0xffff;
1339 Bug("Zero passed to d2b");
1357 #ifndef Sudden_Underflow
1361 *e = (de - Bias - (P-1) << 2) + k;
1362 *bits = 4*P + 8 - k - hi0bits(word0(d) & Frac_mask);
1364 *e = de - Bias - (P-1) + k;
1367 #ifndef Sudden_Underflow
1370 *e = de - Bias - (P-1) + 1 + k;
1372 *bits = 32*i - hi0bits(x[i-1]);
1374 *bits = (i+2)*16 - hi0bits(x[i]);
1386 (a, b) Bigint *a, *b;
1388 (Bigint *a, Bigint *b)
1394 dval(da) = b2d(a, &ka);
1395 dval(db) = b2d(b, &kb);
1397 k = ka - kb + 32*(a->wds - b->wds);
1399 k = ka - kb + 16*(a->wds - b->wds);
1403 word0(da) += (k >> 2)*Exp_msk1;
1409 word0(db) += (k >> 2)*Exp_msk1;
1415 word0(da) += k*Exp_msk1;
1418 word0(db) += k*Exp_msk1;
1421 return dval(da) / dval(db);
1426 1e0, 1e1, 1e2, 1e3, 1e4, 1e5, 1e6, 1e7, 1e8, 1e9,
1427 1e10, 1e11, 1e12, 1e13, 1e14, 1e15, 1e16, 1e17, 1e18, 1e19,
1436 bigtens[] = { 1e16, 1e32, 1e64, 1e128, 1e256 };
1437 static CONST double tinytens[] = { 1e-16, 1e-32, 1e-64, 1e-128,
1438 #ifdef Avoid_Underflow
1439 9007199254740992.*9007199254740992.e-256
1440 /* = 2^106 * 1e-53 */
1445 /* The factor of 2^53 in tinytens[4] helps us avoid setting the underflow */
1446 /* flag unnecessarily. It leads to a song and dance at the end of strtod. */
1447 #define Scale_Bit 0x10
1451 bigtens[] = { 1e16, 1e32, 1e64 };
1452 static CONST double tinytens[] = { 1e-16, 1e-32, 1e-64 };
1455 bigtens[] = { 1e16, 1e32 };
1456 static CONST double tinytens[] = { 1e-16, 1e-32 };
1468 #define NAN_WORD0 0x7ff80000
1478 (sp, t) char **sp, *t;
1480 (CONST char **sp, char *t)
1484 CONST char *s = *sp;
1487 if ((c = *++s) >= 'A' && c <= 'Z')
1500 (rvp, sp) double *rvp; CONST char **sp;
1502 (double *rvp, CONST char **sp)
1507 int havedig, udx0, xshift;
1510 havedig = xshift = 0;
1513 while(c = *(CONST unsigned char*)++s) {
1514 if (c >= '0' && c <= '9')
1516 else if (c >= 'a' && c <= 'f')
1518 else if (c >= 'A' && c <= 'F')
1520 else if (c <= ' ') {
1521 if (udx0 && havedig) {
1527 else if (/*(*/ c == ')' && havedig) {
1532 return; /* invalid form: don't change *sp */
1540 x[0] = (x[0] << 4) | (x[1] >> 28);
1541 x[1] = (x[1] << 4) | c;
1543 if ((x[0] &= 0xfffff) || x[1]) {
1544 word0(*rvp) = Exp_mask | x[0];
1548 #endif /*No_Hex_NaN*/
1549 #endif /* INFNAN_CHECK */
1554 (s00, se) CONST char *s00; char **se;
1556 (CONST char *s00, char **se)
1559 #ifdef Avoid_Underflow
1562 int bb2, bb5, bbe, bd2, bd5, bbbits, bs2, c, dsign,
1563 e, e1, esign, i, j, k, nd, nd0, nf, nz, nz0, sign;
1564 CONST char *s, *s0, *s1;
1565 double aadj, aadj1, adj, rv, rv0;
1568 Bigint *bb = NULL, *bb1, *bd = NULL, *bd0, *bs = NULL, *delta = NULL;
1570 int inexact, oldinexact;
1572 #ifdef Honor_FLT_ROUNDS
1579 sign = nz0 = nz = 0;
1581 for(s = s00;;s++) switch(*s) {
1604 while(*++s == '0') ;
1610 for(nd = nf = 0; (c = *s) >= '0' && c <= '9'; nd++, s++)
1617 s1 = localeconv()->decimal_point;
1638 for(; c == '0'; c = *++s)
1640 if (c > '0' && c <= '9') {
1648 for(; c >= '0' && c <= '9'; c = *++s) {
1653 for(i = 1; i < nz; i++)
1656 else if (nd <= DBL_DIG + 1)
1660 else if (nd <= DBL_DIG + 1)
1668 if (c == 'e' || c == 'E') {
1669 if (!nd && !nz && !nz0) {
1680 if (c >= '0' && c <= '9') {
1683 if (c > '0' && c <= '9') {
1686 while((c = *++s) >= '0' && c <= '9')
1688 if (s - s1 > 8 || L > 19999)
1689 /* Avoid confusion from exponents
1690 * so large that e might overflow.
1692 e = 19999; /* safe for 16 bit ints */
1707 /* Check for Nan and Infinity */
1711 if (match(&s,"nf")) {
1713 if (!match(&s,"inity"))
1715 word0(rv) = 0x7ff00000;
1722 if (match(&s, "an")) {
1723 word0(rv) = NAN_WORD0;
1724 word1(rv) = NAN_WORD1;
1726 if (*s == '(') /*)*/
1732 #endif /* INFNAN_CHECK */
1741 /* Now we have nd0 digits, starting at s0, followed by a
1742 * decimal point, followed by nd-nd0 digits. The number we're
1743 * after is the integer represented by those digits times
1748 k = nd < DBL_DIG + 1 ? nd : DBL_DIG + 1;
1753 oldinexact = get_inexact();
1755 dval(rv) = tens[k - 9] * dval(rv) + z;
1759 #ifndef RND_PRODQUOT
1760 #ifndef Honor_FLT_ROUNDS
1768 if (e <= Ten_pmax) {
1770 goto vax_ovfl_check;
1772 #ifdef Honor_FLT_ROUNDS
1773 /* round correctly FLT_ROUNDS = 2 or 3 */
1779 /* rv = */ rounded_product(dval(rv), tens[e]);
1784 if (e <= Ten_pmax + i) {
1785 /* A fancier test would sometimes let us do
1786 * this for larger i values.
1788 #ifdef Honor_FLT_ROUNDS
1789 /* round correctly FLT_ROUNDS = 2 or 3 */
1796 dval(rv) *= tens[i];
1798 /* VAX exponent range is so narrow we must
1799 * worry about overflow here...
1802 word0(rv) -= P*Exp_msk1;
1803 /* rv = */ rounded_product(dval(rv), tens[e]);
1804 if ((word0(rv) & Exp_mask)
1805 > Exp_msk1*(DBL_MAX_EXP+Bias-1-P))
1807 word0(rv) += P*Exp_msk1;
1809 /* rv = */ rounded_product(dval(rv), tens[e]);
1814 #ifndef Inaccurate_Divide
1815 else if (e >= -Ten_pmax) {
1816 #ifdef Honor_FLT_ROUNDS
1817 /* round correctly FLT_ROUNDS = 2 or 3 */
1823 /* rv = */ rounded_quotient(dval(rv), tens[-e]);
1834 oldinexact = get_inexact();
1836 #ifdef Avoid_Underflow
1839 #ifdef Honor_FLT_ROUNDS
1840 if ((rounding = Flt_Rounds) >= 2) {
1842 rounding = rounding == 2 ? 0 : 2;
1848 #endif /*IEEE_Arith*/
1850 /* Get starting approximation = rv * 10**e1 */
1854 dval(rv) *= tens[i];
1856 if (e1 > DBL_MAX_10_EXP) {
1861 /* Can't trust HUGE_VAL */
1863 #ifdef Honor_FLT_ROUNDS
1865 case 0: /* toward 0 */
1866 case 3: /* toward -infinity */
1871 word0(rv) = Exp_mask;
1874 #else /*Honor_FLT_ROUNDS*/
1875 word0(rv) = Exp_mask;
1877 #endif /*Honor_FLT_ROUNDS*/
1879 /* set overflow bit */
1881 dval(rv0) *= dval(rv0);
1883 #else /*IEEE_Arith*/
1886 #endif /*IEEE_Arith*/
1892 for(j = 0; e1 > 1; j++, e1 >>= 1)
1894 dval(rv) *= bigtens[j];
1895 /* The last multiplication could overflow. */
1896 word0(rv) -= P*Exp_msk1;
1897 dval(rv) *= bigtens[j];
1898 if ((z = word0(rv) & Exp_mask)
1899 > Exp_msk1*(DBL_MAX_EXP+Bias-P))
1901 if (z > Exp_msk1*(DBL_MAX_EXP+Bias-1-P)) {
1902 /* set to largest number */
1903 /* (Can't trust DBL_MAX) */
1908 word0(rv) += P*Exp_msk1;
1914 dval(rv) /= tens[i];
1916 if (e1 >= 1 << n_bigtens)
1918 #ifdef Avoid_Underflow
1921 for(j = 0; e1 > 0; j++, e1 >>= 1)
1923 dval(rv) *= tinytens[j];
1924 if (scale && (j = 2*P + 1 - ((word0(rv) & Exp_mask)
1925 >> Exp_shift)) > 0) {
1926 /* scaled rv is denormal; zap j low bits */
1930 word0(rv) = (P+2)*Exp_msk1;
1932 word0(rv) &= 0xffffffff << (j-32);
1935 word1(rv) &= 0xffffffff << j;
1938 for(j = 0; e1 > 1; j++, e1 >>= 1)
1940 dval(rv) *= tinytens[j];
1941 /* The last multiplication could underflow. */
1942 dval(rv0) = dval(rv);
1943 dval(rv) *= tinytens[j];
1945 dval(rv) = 2.*dval(rv0);
1946 dval(rv) *= tinytens[j];
1958 #ifndef Avoid_Underflow
1961 /* The refinement below will clean
1962 * this approximation up.
1969 /* Now the hard part -- adjusting rv to the correct value.*/
1971 /* Put digits into bd: true value = bd * 10^e */
1973 bd0 = s2b(s0, nd0, nd, y);
1976 bd = Balloc(bd0->k);
1978 bb = d2b(dval(rv), &bbe, &bbbits); /* rv = bb * 2^bbe */
1994 #ifdef Honor_FLT_ROUNDS
1998 #ifdef Avoid_Underflow
2000 i = j + bbbits - 1; /* logb(rv) */
2001 if (i < Emin) /* denormal */
2005 #else /*Avoid_Underflow*/
2006 #ifdef Sudden_Underflow
2008 j = 1 + 4*P - 3 - bbbits + ((bbe + bbbits - 1) & 3);
2012 #else /*Sudden_Underflow*/
2014 i = j + bbbits - 1; /* logb(rv) */
2015 if (i < Emin) /* denormal */
2019 #endif /*Sudden_Underflow*/
2020 #endif /*Avoid_Underflow*/
2023 #ifdef Avoid_Underflow
2026 i = bb2 < bd2 ? bb2 : bd2;
2035 bs = pow5mult(bs, bb5);
2041 bb = lshift(bb, bb2);
2043 bd = pow5mult(bd, bd5);
2045 bd = lshift(bd, bd2);
2047 bs = lshift(bs, bs2);
2048 delta = diff(bb, bd);
2049 dsign = delta->sign;
2052 #ifdef Honor_FLT_ROUNDS
2053 if (rounding != 1) {
2055 /* Error is less than an ulp */
2056 if (!delta->x[0] && delta->wds <= 1) {
2072 && !(word0(rv) & Frac_mask)) {
2073 y = word0(rv) & Exp_mask;
2074 #ifdef Avoid_Underflow
2075 if (!scale || y > 2*P*Exp_msk1)
2080 delta = lshift(delta,Log2P);
2081 if (cmp(delta, bs) <= 0)
2086 #ifdef Avoid_Underflow
2087 if (scale && (y = word0(rv) & Exp_mask)
2089 word0(adj) += (2*P+1)*Exp_msk1 - y;
2091 #ifdef Sudden_Underflow
2092 if ((word0(rv) & Exp_mask) <=
2094 word0(rv) += P*Exp_msk1;
2095 dval(rv) += adj*ulp(dval(rv));
2096 word0(rv) -= P*Exp_msk1;
2099 #endif /*Sudden_Underflow*/
2100 #endif /*Avoid_Underflow*/
2101 dval(rv) += adj*ulp(dval(rv));
2105 adj = ratio(delta, bs);
2108 if (adj <= 0x7ffffffe) {
2109 /* adj = rounding ? ceil(adj) : floor(adj); */
2112 if (!((rounding>>1) ^ dsign))
2117 #ifdef Avoid_Underflow
2118 if (scale && (y = word0(rv) & Exp_mask) <= 2*P*Exp_msk1)
2119 word0(adj) += (2*P+1)*Exp_msk1 - y;
2121 #ifdef Sudden_Underflow
2122 if ((word0(rv) & Exp_mask) <= P*Exp_msk1) {
2123 word0(rv) += P*Exp_msk1;
2124 adj *= ulp(dval(rv));
2129 word0(rv) -= P*Exp_msk1;
2132 #endif /*Sudden_Underflow*/
2133 #endif /*Avoid_Underflow*/
2134 adj *= ulp(dval(rv));
2141 #endif /*Honor_FLT_ROUNDS*/
2144 /* Error is less than half an ulp -- check for
2145 * special case of mantissa a power of two.
2147 if (dsign || word1(rv) || word0(rv) & Bndry_mask
2149 #ifdef Avoid_Underflow
2150 || (word0(rv) & Exp_mask) <= (2*P+1)*Exp_msk1
2152 || (word0(rv) & Exp_mask) <= Exp_msk1
2157 if (!delta->x[0] && delta->wds <= 1)
2162 if (!delta->x[0] && delta->wds <= 1) {
2169 delta = lshift(delta,Log2P);
2170 if (cmp(delta, bs) > 0)
2175 /* exactly half-way between */
2177 if ((word0(rv) & Bndry_mask1) == Bndry_mask1
2179 #ifdef Avoid_Underflow
2180 (scale && (y = word0(rv) & Exp_mask) <= 2*P*Exp_msk1)
2181 ? (0xffffffff & (0xffffffff << (2*P+1-(y>>Exp_shift)))) :
2184 /*boundary case -- increment exponent*/
2185 word0(rv) = (word0(rv) & Exp_mask)
2192 #ifdef Avoid_Underflow
2198 else if (!(word0(rv) & Bndry_mask) && !word1(rv)) {
2200 /* boundary case -- decrement exponent */
2201 #ifdef Sudden_Underflow /*{{*/
2202 L = word0(rv) & Exp_mask;
2206 #ifdef Avoid_Underflow
2207 if (L <= (scale ? (2*P+1)*Exp_msk1 : Exp_msk1))
2210 #endif /*Avoid_Underflow*/
2214 #else /*Sudden_Underflow}{*/
2215 #ifdef Avoid_Underflow
2217 L = word0(rv) & Exp_mask;
2218 if (L <= (2*P+1)*Exp_msk1) {
2219 if (L > (P+2)*Exp_msk1)
2220 /* round even ==> */
2223 /* rv = smallest denormal */
2227 #endif /*Avoid_Underflow*/
2228 L = (word0(rv) & Exp_mask) - Exp_msk1;
2229 #endif /*Sudden_Underflow}}*/
2230 word0(rv) = L | Bndry_mask1;
2231 word1(rv) = 0xffffffff;
2238 #ifndef ROUND_BIASED
2239 if (!(word1(rv) & LSB))
2243 dval(rv) += ulp(dval(rv));
2244 #ifndef ROUND_BIASED
2246 dval(rv) -= ulp(dval(rv));
2247 #ifndef Sudden_Underflow
2252 #ifdef Avoid_Underflow
2258 if ((aadj = ratio(delta, bs)) <= 2.) {
2261 else if (word1(rv) || word0(rv) & Bndry_mask) {
2262 #ifndef Sudden_Underflow
2263 if (word1(rv) == Tiny1 && !word0(rv))
2270 /* special case -- power of FLT_RADIX to be */
2271 /* rounded down... */
2273 if (aadj < 2./FLT_RADIX)
2274 aadj = 1./FLT_RADIX;
2282 aadj1 = dsign ? aadj : -aadj;
2283 #ifdef Check_FLT_ROUNDS
2285 case 2: /* towards +infinity */
2288 case 0: /* towards 0 */
2289 case 3: /* towards -infinity */
2293 if (Flt_Rounds == 0)
2295 #endif /*Check_FLT_ROUNDS*/
2297 y = word0(rv) & Exp_mask;
2299 /* Check for overflow */
2301 if (y == Exp_msk1*(DBL_MAX_EXP+Bias-1)) {
2302 dval(rv0) = dval(rv);
2303 word0(rv) -= P*Exp_msk1;
2304 adj = aadj1 * ulp(dval(rv));
2306 if ((word0(rv) & Exp_mask) >=
2307 Exp_msk1*(DBL_MAX_EXP+Bias-P)) {
2308 if (word0(rv0) == Big0 && word1(rv0) == Big1)
2315 word0(rv) += P*Exp_msk1;
2318 #ifdef Avoid_Underflow
2319 if (scale && y <= 2*P*Exp_msk1) {
2320 if (aadj <= 0x7fffffff) {
2321 if ((z = aadj) <= 0)
2324 aadj1 = dsign ? aadj : -aadj;
2326 word0(aadj1) += (2*P+1)*Exp_msk1 - y;
2328 adj = aadj1 * ulp(dval(rv));
2331 #ifdef Sudden_Underflow
2332 if ((word0(rv) & Exp_mask) <= P*Exp_msk1) {
2333 dval(rv0) = dval(rv);
2334 word0(rv) += P*Exp_msk1;
2335 adj = aadj1 * ulp(dval(rv));
2338 if ((word0(rv) & Exp_mask) < P*Exp_msk1)
2340 if ((word0(rv) & Exp_mask) <= P*Exp_msk1)
2343 if (word0(rv0) == Tiny0
2344 && word1(rv0) == Tiny1)
2351 word0(rv) -= P*Exp_msk1;
2354 adj = aadj1 * ulp(dval(rv));
2357 #else /*Sudden_Underflow*/
2358 /* Compute adj so that the IEEE rounding rules will
2359 * correctly round rv + adj in some half-way cases.
2360 * If rv * ulp(rv) is denormalized (i.e.,
2361 * y <= (P-1)*Exp_msk1), we must adjust aadj to avoid
2362 * trouble from bits lost to denormalization;
2363 * example: 1.2e-307 .
2365 if (y <= (P-1)*Exp_msk1 && aadj > 1.) {
2366 aadj1 = (double)(int)(aadj + 0.5);
2370 adj = aadj1 * ulp(dval(rv));
2372 #endif /*Sudden_Underflow*/
2373 #endif /*Avoid_Underflow*/
2375 z = word0(rv) & Exp_mask;
2377 #ifdef Avoid_Underflow
2381 /* Can we stop now? */
2384 /* The tolerances below are conservative. */
2385 if (dsign || word1(rv) || word0(rv) & Bndry_mask) {
2386 if (aadj < .4999999 || aadj > .5000001)
2389 else if (aadj < .4999999/FLT_RADIX)
2402 word0(rv0) = Exp_1 + (70 << Exp_shift);
2407 else if (!oldinexact)
2410 #ifdef Avoid_Underflow
2412 word0(rv0) = Exp_1 - 2*P*Exp_msk1;
2414 dval(rv) *= dval(rv0);
2416 /* try to avoid the bug of testing an 8087 register value */
2417 if (word0(rv) == 0 && word1(rv) == 0)
2421 #endif /* Avoid_Underflow */
2423 if (inexact && !(word0(rv) & Exp_mask)) {
2424 /* set underflow bit */
2426 dval(rv0) *= dval(rv0);
2438 return sign ? -dval(rv) : dval(rv);
2444 (b, S) Bigint *b, *S;
2446 (Bigint *b, Bigint *S)
2450 ULong *bx, *bxe, q, *sx, *sxe;
2452 ULLong borrow, carry, y, ys;
2454 ULong borrow, carry, y, ys;
2462 /*debug*/ if (b->wds > n)
2463 /*debug*/ Bug("oversize b in quorem");
2471 q = *bxe / (*sxe + 1); /* ensure q <= true quotient */
2473 /*debug*/ if (q > 9)
2474 /*debug*/ Bug("oversized quotient in quorem");
2481 ys = *sx++ * (ULLong)q + carry;
2483 y = *bx - (ys & FFFFFFFF) - borrow;
2484 borrow = y >> 32 & (ULong)1;
2485 *bx++ = y & FFFFFFFF;
2489 ys = (si & 0xffff) * q + carry;
2490 zs = (si >> 16) * q + (ys >> 16);
2492 y = (*bx & 0xffff) - (ys & 0xffff) - borrow;
2493 borrow = (y & 0x10000) >> 16;
2494 z = (*bx >> 16) - (zs & 0xffff) - borrow;
2495 borrow = (z & 0x10000) >> 16;
2498 ys = *sx++ * q + carry;
2500 y = *bx - (ys & 0xffff) - borrow;
2501 borrow = (y & 0x10000) >> 16;
2509 while(--bxe > bx && !*bxe)
2514 if (cmp(b, S) >= 0) {
2524 y = *bx - (ys & FFFFFFFF) - borrow;
2525 borrow = y >> 32 & (ULong)1;
2526 *bx++ = y & FFFFFFFF;
2530 ys = (si & 0xffff) + carry;
2531 zs = (si >> 16) + (ys >> 16);
2533 y = (*bx & 0xffff) - (ys & 0xffff) - borrow;
2534 borrow = (y & 0x10000) >> 16;
2535 z = (*bx >> 16) - (zs & 0xffff) - borrow;
2536 borrow = (z & 0x10000) >> 16;
2541 y = *bx - (ys & 0xffff) - borrow;
2542 borrow = (y & 0x10000) >> 16;
2551 while(--bxe > bx && !*bxe)
2559 #ifndef MULTIPLE_THREADS
2560 static char *dtoa_result;
2574 sizeof(Bigint) - sizeof(ULong) - sizeof(int) + j <= i;
2577 r = (int*)Balloc(k);
2580 #ifndef MULTIPLE_THREADS
2588 nrv_alloc(s, rve, n) char *s, **rve; int n;
2590 nrv_alloc(char *s, char **rve, int n)
2595 t = rv = rv_alloc(n);
2596 while((*t = *s++)) t++;
2602 /* freedtoa(s) must be used to free values s returned by dtoa
2603 * when MULTIPLE_THREADS is #defined. It should be used in all cases,
2604 * but for consistency with earlier versions of dtoa, it is optional
2605 * when MULTIPLE_THREADS is not defined.
2608 static void freedtoa (char *s);
2612 freedtoa(s) char *s;
2617 Bigint *b = (Bigint *)((int *)s - 1);
2618 b->maxwds = 1 << (b->k = *(int*)b);
2620 #ifndef MULTIPLE_THREADS
2621 if (s == dtoa_result)
2627 /* dtoa for IEEE arithmetic (dmg): convert double to ASCII string.
2629 * Inspired by "How to Print Floating-Point Numbers Accurately" by
2630 * Guy L. Steele, Jr. and Jon L. White [Proc. ACM SIGPLAN '90, pp. 112-126].
2633 * 1. Rather than iterating, we use a simple numeric overestimate
2634 * to determine k = floor(log10(d)). We scale relevant
2635 * quantities using O(log2(k)) rather than O(k) multiplications.
2636 * 2. For some modes > 2 (corresponding to ecvt and fcvt), we don't
2637 * try to generate digits strictly left to right. Instead, we
2638 * compute with fewer bits and propagate the carry if necessary
2639 * when rounding the final digit up. This is often faster.
2640 * 3. Under the assumption that input will be rounded nearest,
2641 * mode 0 renders 1e23 as 1e23 rather than 9.999999999999999e22.
2642 * That is, we allow equality in stopping tests when the
2643 * round-nearest rule will give the same floating-point value
2644 * as would satisfaction of the stopping test with strict
2646 * 4. We remove common factors of powers of 2 from relevant
2648 * 5. When converting floating-point integers less than 1e16,
2649 * we use floating-point arithmetic rather than resorting
2650 * to multiple-precision integers.
2651 * 6. When asked to produce fewer than 15 digits, we first try
2652 * to get by with floating-point arithmetic; we resort to
2653 * multiple-precision integer arithmetic only if we cannot
2654 * guarantee that the floating-point calculation has given
2655 * the correctly rounded result. For k requested digits and
2656 * "uniformly" distributed input, the probability is
2657 * something like 10^(k-15) that we must resort to the Long
2664 (d, mode, ndigits, decpt, sign, rve)
2665 double d; int mode, ndigits, *decpt, *sign; char **rve;
2667 (double d, int mode, int ndigits, int *decpt, int *sign, char **rve)
2670 /* Arguments ndigits, decpt, sign are similar to those
2671 of ecvt and fcvt; trailing zeros are suppressed from
2672 the returned string. If not null, *rve is set to point
2673 to the end of the return value. If d is +-Infinity or NaN,
2674 then *decpt is set to 9999.
2677 0 ==> shortest string that yields d when read in
2678 and rounded to nearest.
2679 1 ==> like 0, but with Steele & White stopping rule;
2680 e.g. with IEEE P754 arithmetic , mode 0 gives
2681 1e23 whereas mode 1 gives 9.999999999999999e22.
2682 2 ==> max(1,ndigits) significant digits. This gives a
2683 return value similar to that of ecvt, except
2684 that trailing zeros are suppressed.
2685 3 ==> through ndigits past the decimal point. This
2686 gives a return value similar to that from fcvt,
2687 except that trailing zeros are suppressed, and
2688 ndigits can be negative.
2689 4,5 ==> similar to 2 and 3, respectively, but (in
2690 round-nearest mode) with the tests of mode 0 to
2691 possibly return a shorter string that rounds to d.
2692 With IEEE arithmetic and compilation with
2693 -DHonor_FLT_ROUNDS, modes 4 and 5 behave the same
2694 as modes 2 and 3 when FLT_ROUNDS != 1.
2695 6-9 ==> Debugging modes similar to mode - 4: don't try
2696 fast floating-point estimate (if applicable).
2698 Values of mode other than 0-9 are treated as mode 0.
2700 Sufficient space is allocated to the return value
2701 to hold the suppressed trailing zeros.
2704 int bbits, b2, b5, be, dig, i, ieps, ilim, ilim0, ilim1,
2705 j, j1, k, k0, k_check, leftright, m2, m5, s2, s5,
2706 spec_case, try_quick;
2708 #ifndef Sudden_Underflow
2712 Bigint *b, *b1, *delta, *mlo, *mhi, *S;
2715 #ifdef Honor_FLT_ROUNDS
2719 int inexact, oldinexact;
2722 #ifndef MULTIPLE_THREADS
2724 freedtoa(dtoa_result);
2729 if (word0(d) & Sign_bit) {
2730 /* set sign for everything, including 0's and NaNs */
2732 word0(d) &= ~Sign_bit; /* clear sign bit */
2737 #if defined(IEEE_Arith) + defined(VAX)
2739 if ((word0(d) & Exp_mask) == Exp_mask)
2741 if (word0(d) == 0x8000)
2744 /* Infinity or NaN */
2747 if (!word1(d) && !(word0(d) & 0xfffff))
2748 return nrv_alloc("Infinity", rve, 8);
2750 return nrv_alloc("NaN", rve, 3);
2754 dval(d) += 0; /* normalize */
2758 return nrv_alloc("0", rve, 1);
2762 try_quick = oldinexact = get_inexact();
2765 #ifdef Honor_FLT_ROUNDS
2766 if ((rounding = Flt_Rounds) >= 2) {
2768 rounding = rounding == 2 ? 0 : 2;
2775 b = d2b(dval(d), &be, &bbits);
2776 #ifdef Sudden_Underflow
2777 i = (int)(word0(d) >> Exp_shift1 & (Exp_mask>>Exp_shift1));
2779 if (i = (int)(word0(d) >> Exp_shift1 & (Exp_mask>>Exp_shift1))) {
2782 word0(d2) &= Frac_mask1;
2783 word0(d2) |= Exp_11;
2785 if (j = 11 - hi0bits(word0(d2) & Frac_mask))
2789 /* log(x) ~=~ log(1.5) + (x-1.5)/1.5
2790 * log10(x) = log(x) / log(10)
2791 * ~=~ log(1.5)/log(10) + (x-1.5)/(1.5*log(10))
2792 * log10(d) = (i-Bias)*log(2)/log(10) + log10(d2)
2794 * This suggests computing an approximation k to log10(d) by
2796 * k = (i - Bias)*0.301029995663981
2797 * + ( (d2-1.5)*0.289529654602168 + 0.176091259055681 );
2799 * We want k to be too large rather than too small.
2800 * The error in the first-order Taylor series approximation
2801 * is in our favor, so we just round up the constant enough
2802 * to compensate for any error in the multiplication of
2803 * (i - Bias) by 0.301029995663981; since |i - Bias| <= 1077,
2804 * and 1077 * 0.30103 * 2^-52 ~=~ 7.2e-14,
2805 * adding 1e-13 to the constant term more than suffices.
2806 * Hence we adjust the constant term to 0.1760912590558.
2807 * (We could get a more accurate k by invoking log10,
2808 * but this is probably not worthwhile.)
2816 #ifndef Sudden_Underflow
2820 /* d is denormalized */
2822 i = bbits + be + (Bias + (P-1) - 1);
2823 x = i > 32 ? word0(d) << 64 - i | word1(d) >> i - 32
2824 : word1(d) << 32 - i;
2826 word0(d2) -= 31*Exp_msk1; /* adjust exponent */
2827 i -= (Bias + (P-1) - 1) + 1;
2831 ds = (dval(d2)-1.5)*0.289529654602168 + 0.1760912590558 + i*0.301029995663981;
2833 if (ds < 0. && ds != k)
2834 k--; /* want k = floor(ds) */
2836 if (k >= 0 && k <= Ten_pmax) {
2837 if (dval(d) < tens[k])
2860 if (mode < 0 || mode > 9)
2864 #ifdef Check_FLT_ROUNDS
2865 try_quick = Rounding == 1;
2869 #endif /*SET_INEXACT*/
2889 ilim = ilim1 = i = ndigits;
2895 i = ndigits + k + 1;
2901 s = s0 = rv_alloc(i);
2903 #ifdef Honor_FLT_ROUNDS
2904 if (mode > 1 && rounding != 1)
2908 if (ilim >= 0 && ilim <= Quick_max && try_quick) {
2910 /* Try to get by with floating-point arithmetic. */
2916 ieps = 2; /* conservative */
2921 /* prevent overflows */
2923 dval(d) /= bigtens[n_bigtens-1];
2926 for(; j; j >>= 1, i++)
2934 dval(d) *= tens[j1 & 0xf];
2935 for(j = j1 >> 4; j; j >>= 1, i++)
2938 dval(d) *= bigtens[i];
2941 if (k_check && dval(d) < 1. && ilim > 0) {
2949 dval(eps) = ieps*dval(d) + 7.;
2950 word0(eps) -= (P-1)*Exp_msk1;
2954 if (dval(d) > dval(eps))
2956 if (dval(d) < -dval(eps))
2960 #ifndef No_leftright
2962 /* Use Steele & White method of only
2963 * generating digits needed.
2965 dval(eps) = 0.5/tens[ilim-1] - dval(eps);
2969 *s++ = '0' + (int)L;
2970 if (dval(d) < dval(eps))
2972 if (1. - dval(d) < dval(eps))
2982 /* Generate ilim digits, then fix them up. */
2983 dval(eps) *= tens[ilim-1];
2984 for(i = 1;; i++, dval(d) *= 10.) {
2985 L = (Long)(dval(d));
2986 if (!(dval(d) -= L))
2988 *s++ = '0' + (int)L;
2990 if (dval(d) > 0.5 + dval(eps))
2992 else if (dval(d) < 0.5 - dval(eps)) {
3000 #ifndef No_leftright
3010 /* Do we have a "small" integer? */
3012 if (be >= 0 && k <= Int_max) {
3015 if (ndigits < 0 && ilim <= 0) {
3017 if (ilim < 0 || dval(d) <= 5*ds)
3021 for(i = 1;; i++, dval(d) *= 10.) {
3022 L = (Long)(dval(d) / ds);
3024 #ifdef Check_FLT_ROUNDS
3025 /* If FLT_ROUNDS == 2, L will usually be high by 1 */
3031 *s++ = '0' + (int)L;
3039 #ifdef Honor_FLT_ROUNDS
3043 case 2: goto bump_up;
3047 if (dval(d) > ds || dval(d) == ds && L & 1) {
3068 #ifndef Sudden_Underflow
3069 denorm ? be + (Bias + (P-1) - 1 + 1) :
3072 1 + 4*P - 3 - bbits + ((bbits + be - 1) & 3);
3080 if (m2 > 0 && s2 > 0) {
3081 i = m2 < s2 ? m2 : s2;
3089 mhi = pow5mult(mhi, m5);
3098 b = pow5mult(b, b5);
3102 S = pow5mult(S, s5);
3104 /* Check for special case that d is a normalized power of 2. */
3107 if ((mode < 2 || leftright)
3108 #ifdef Honor_FLT_ROUNDS
3112 if (!word1(d) && !(word0(d) & Bndry_mask)
3113 #ifndef Sudden_Underflow
3114 && word0(d) & (Exp_mask & ~Exp_msk1)
3117 /* The special case */
3124 /* Arrange for convenient computation of quotients:
3125 * shift left if necessary so divisor has 4 leading 0 bits.
3127 * Perhaps we should just compute leading 28 bits of S once
3128 * and for all and pass them and a shift to quorem, so it
3129 * can do shifts and ors to compute the numerator for q.
3132 if (i = ((s5 ? 32 - hi0bits(S->x[S->wds-1]) : 1) + s2) & 0x1f)
3135 if (i = ((s5 ? 32 - hi0bits(S->x[S->wds-1]) : 1) + s2) & 0xf)
3157 b = multadd(b, 10, 0); /* we botched the k estimate */
3159 mhi = multadd(mhi, 10, 0);
3163 if (ilim <= 0 && (mode == 3 || mode == 5)) {
3164 if (ilim < 0 || cmp(b,S = multadd(S,5,0)) <= 0) {
3165 /* no digits, fcvt style */
3177 mhi = lshift(mhi, m2);
3179 /* Compute mlo -- check for special case
3180 * that d is a normalized power of 2.
3185 mhi = Balloc(mhi->k);
3187 mhi = lshift(mhi, Log2P);
3191 dig = quorem(b,S) + '0';
3192 /* Do we yet have the shortest decimal string
3193 * that will round to d?
3196 delta = diff(S, mhi);
3197 j1 = delta->sign ? 1 : cmp(b, delta);
3199 #ifndef ROUND_BIASED
3200 if (j1 == 0 && mode != 1 && !(word1(d) & 1)
3201 #ifdef Honor_FLT_ROUNDS
3210 else if (!b->x[0] && b->wds <= 1)
3217 if (j < 0 || j == 0 && mode != 1
3218 #ifndef ROUND_BIASED
3222 if (!b->x[0] && b->wds <= 1) {
3228 #ifdef Honor_FLT_ROUNDS
3231 case 0: goto accept_dig;
3232 case 2: goto keep_dig;
3234 #endif /*Honor_FLT_ROUNDS*/
3238 if ((j1 > 0 || j1 == 0 && dig & 1)
3247 #ifdef Honor_FLT_ROUNDS
3251 if (dig == '9') { /* possible if i == 1 */
3259 #ifdef Honor_FLT_ROUNDS
3265 b = multadd(b, 10, 0);
3267 mlo = mhi = multadd(mhi, 10, 0);
3269 mlo = multadd(mlo, 10, 0);
3270 mhi = multadd(mhi, 10, 0);
3276 *s++ = dig = quorem(b,S) + '0';
3277 if (!b->x[0] && b->wds <= 1) {
3285 b = multadd(b, 10, 0);
3288 /* Round off last digit */
3290 #ifdef Honor_FLT_ROUNDS
3292 case 0: goto trimzeros;
3293 case 2: goto roundoff;
3298 if (j > 0 || j == 0 && dig & 1) {
3316 if (mlo && mlo != mhi)
3324 word0(d) = Exp_1 + (70 << Exp_shift);
3329 else if (!oldinexact)