summaryrefslogtreecommitdiff
path: root/src/exp10m1.c
blob: d5a7663b43b1a9e24765720c29eac54c5b304002 (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
/* mpfr_exp10m1 -- Compute 10^x-1

Copyright 2001-2021 Free Software Foundation, Inc.
Contributed by the AriC and Caramba projects, INRIA.

This file is part of the GNU MPFR Library.

The GNU MPFR Library is free software; you can redistribute it and/or modify
it under the terms of the GNU Lesser General Public License as published by
the Free Software Foundation; either version 3 of the License, or (at your
option) any later version.

The GNU MPFR Library is distributed in the hope that it will be useful, but
WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU Lesser General Public
License for more details.

You should have received a copy of the GNU Lesser General Public License
along with the GNU MPFR Library; see the file COPYING.LESSER.  If not, see
https://www.gnu.org/licenses/ or write to the Free Software Foundation, Inc.,
51 Franklin St, Fifth Floor, Boston, MA 02110-1301, USA. */

#define MPFR_NEED_LONGLONG_H
#include "mpfr-impl.h"

/* The computation of exp10m1 is done by expm1(x) = 10^x-1 */

/* In case x is small in absolute value, 10^x - 1 ~ x*log(10).
   If this is enough to deduce correct rounding, put in the auxiliary variable
   t the approximation that will be rounded to get y, and return non-zero.
   If we put 0 in t, it means underflow.
   Otherwise return 0. */
static int
mpfr_exp10m1_small (mpfr_ptr y, mpfr_srcptr x , mpfr_rnd_t rnd_mode,
                    mpfr_ptr t)
{
  mpfr_prec_t prec;
  mpfr_exp_t e;

  /* for |x| < 0.25, we have |10^x-1-x*log(10)| < 4*x^2 */
  if (MPFR_EXP(x) > -2)
    return 0;
  /* now EXP(x) <= -2, thus x < 0.25 */
  prec = MPFR_PREC(t);
  mpfr_log_ui (t, 10, MPFR_RNDN);
  /* t = log(10)*(1 + theta) with |theta| <= 2^(-prec) */
  mpfr_mul (t, t, x, MPFR_RNDN);
  /* no underflow can occur, since log(10) > 1 */
  /* t = x*log(10)*(1 + theta)^2 with |theta| <= 2^(-prec) */
  /* |t - x*log(10)| <= ((1 + theta)^2 - 1) * |t| <= 3*2^(-prec)*|t| */
  /* |t - x*log(10)| < 3*2^(EXP(t)-prec) */
  e = 2 * MPFR_GET_EXP (x) + 2 + prec - MPFR_GET_EXP(t);
  /* |4*x^2| < 2^e*2^(EXP(t)-prec) thus
     |t - exp10m1(x)| < (3+2^e)*2^(EXP(t)-prec) */
  e = (e <= 1) ? 2 + (e == 1) : e + 1;
  /* now |t - exp10m1(x)| < 2^e*2^(EXP(t)-prec) */
  return MPFR_CAN_ROUND (t, prec - e, MPFR_PREC(y), rnd_mode);
}

int
mpfr_exp10m1 (mpfr_ptr y, mpfr_srcptr x , mpfr_rnd_t rnd_mode)
{
  int inexact, nloop;
  mpfr_t t;
  mpfr_prec_t Ny = MPFR_PREC(y);   /* target precision */
  mpfr_prec_t Nt;                  /* working precision */
  mpfr_exp_t err, exp_te;          /* error */
  MPFR_ZIV_DECL (loop);
  MPFR_SAVE_EXPO_DECL (expo);

  MPFR_LOG_FUNC
    (("x[%Pu]=%.*Rg rnd=%d", mpfr_get_prec (x), mpfr_log_prec, x, rnd_mode),
     ("y[%Pu]=%.*Rg inexact=%d", mpfr_get_prec (y), mpfr_log_prec, y,
      inexact));

  if (MPFR_IS_SINGULAR (x))
    return mpfr_expm1 (y, x, rnd_mode); /* singular cases are identical */

  MPFR_ASSERTN(!MPFR_IS_ZERO(x));

  MPFR_SAVE_EXPO_MARK (expo);

  /* Check case where result is -1 or nextabove(-1) because x is a huge
     negative number. */
  if (MPFR_IS_NEG(x) && mpfr_cmpabs_ui (x, 2 + (MPFR_PREC(y) - 1) / 3) > 0)
    {
      MPFR_SAVE_EXPO_UPDATE_FLAGS (expo, MPFR_FLAGS_INEXACT);
      /* 1/2*ulp(-1) = 2^(-PREC(y)):
         since |x| >= PREC(y)/3 + 1, then 3*|x| >= PREC(y) + 3,
         thus 10^x < 8^x <= 2^(-PREC(y)-3) <= 1/2*ulp(-1), thus the
         result is -1 for RNDA,RNDD,RNDN, and nextabove(-1) for RNDZ,RNDU */
      mpfr_set_si (y, -1, MPFR_RNDZ);
      if (!MPFR_IS_LIKE_RNDZ(rnd_mode,1))
        inexact = -1;
      else
        {
          mpfr_nextabove (y);
          inexact = 1;
        }
      goto end;
    }

  Nt = Ny + MPFR_INT_CEIL_LOG2 (Ny) + 6;

  mpfr_init2 (t, Nt);

  MPFR_ZIV_INIT (loop, Nt);
  for (nloop = 0;; nloop++)
    {
      int inex1;

      MPFR_BLOCK_DECL (flags);

      /* 10^x may overflow and underflow */
      MPFR_BLOCK (flags, inex1 = mpfr_exp10 (t, x, MPFR_RNDN));

      if (MPFR_OVERFLOW (flags)) /* overflow case */
        {
          inexact = mpfr_overflow (y, rnd_mode, MPFR_SIGN_POS);
          MPFR_SAVE_EXPO_UPDATE_FLAGS (expo, MPFR_FLAGS_OVERFLOW);
          goto clear;
        }

      /* integer case */
      if (inex1 == 0)
        {
          inexact = mpfr_sub_ui (y, t, 1, rnd_mode);
          goto clear;
        }

      /* The case of underflow in 10^x (huge negative x)
         was already detected before Ziv's loop. */
      MPFR_ASSERTD(!MPFR_UNDERFLOW (flags));

      MPFR_ASSERTN(!MPFR_IS_ZERO(t));
      exp_te = MPFR_GET_EXP (t);
      mpfr_sub_ui (t, t, 1, MPFR_RNDN);   /* 10^x-1 */

      /* error estimate */
      /* err = __gmpfr_ceil_log2(1+pow(2,MPFR_EXP(te)-MPFR_EXP(t))) */
      if (!MPFR_IS_ZERO(t))
        {
          err = MAX (exp_te - MPFR_GET_EXP (t), 0) + 1;
          /* if inex1=0, this means that t=o(10^x) is exact, thus the correct
             rounding is simply o(t-1) */
          if (inex1 == 0 ||
              MPFR_LIKELY (MPFR_CAN_ROUND (t, Nt - err, Ny, rnd_mode)))
            break;
        }

      /* check small case: we need to do it at each step of Ziv's loop,
         since the multiplication x*log(10) might not enable correct
         rounding at the first loop */
      if (mpfr_exp10m1_small (y, x, rnd_mode, t))
        {
          if (MPFR_IS_ZERO(t)) /* underflow */
            {
              mpfr_clear (t);
              MPFR_SAVE_EXPO_FREE (expo);
              return mpfr_underflow (y, (rnd_mode == MPFR_RNDN) ? MPFR_RNDZ
                                     : rnd_mode, 1);
            }
          break;
        }

      /* increase the precision */
      MPFR_ZIV_NEXT (loop, Nt);
      mpfr_set_prec (t, Nt);
    }

  inexact = mpfr_set (y, t, rnd_mode);
 clear:
  MPFR_ZIV_FREE (loop);
  mpfr_clear (t);

 end:
  MPFR_SAVE_EXPO_FREE (expo);
  return mpfr_check_range (y, inexact, rnd_mode);
}