summaryrefslogtreecommitdiff
path: root/src/log2p1.c
blob: 197e2876b4cfee43082e88750332463b26542608 (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
/* mpfr_log2p1 -- Compute log2(1+x)

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 /* needed for MPFR_INT_CEIL_LOG2 */
#include "mpfr-impl.h"

#define ULSIZE (sizeof (unsigned long) * CHAR_BIT)

/* return non-zero if log2(1+x) is exactly representable in infinite precision,
   and in such case the returned value is k such that 1+x = 2^k (the case k=0
   cannot happen since we assume x<>0) */
static mpfr_exp_t
mpfr_log2p1_isexact (mpfr_srcptr x)
{
  /* log2(1+x) is exactly representable when 1+x is a power of two,
     we thus simply compute 1+x with 1-bit precision and check whether
     the addition is exact. This routine is called with extended exponent
     range, thus no need to extend it. */
  mpfr_t t;
  int inex;
  mpfr_exp_t e;

  mpfr_init2 (t, 1);
  inex = mpfr_add_ui (t, x, 1, MPFR_RNDZ);
  e = MPFR_GET_EXP (t);
  mpfr_clear (t);
  return inex == 0 ? e - 1 : 0;
}

/* in case x=2^k and we can decide of the correct rounding,
   put the correctly-rounded value in y and return the corresponding
   ternary value (which is necessarily non-zero),
   otherwise return 0 */
static int
mpfr_log2p1_special (mpfr_ptr y, mpfr_srcptr x, mpfr_rnd_t rnd_mode)
{
  mpfr_exp_t expx = MPFR_GET_EXP(x);
  mpfr_exp_t k = expx - 1, expk;
  mpfr_prec_t prec;
  mpfr_t t;
  int inex;

  if (k <= 0 || mpfr_cmp_si_2exp (x, 1, k) != 0)
    return 0;
  /* k < log2(1+x) < k + 1/x/log(2) < k + 2/x */
  expk = MPFR_INT_CEIL_LOG2(k); /* exponent of k */
  /* 2/x < 2^(2-EXP(x)) thus if 2-EXP(x) < expk - PREC(y) - 1,
     we have 2/x < 1/4*ulp(k) and we can decide the correct rounding */
  if (2 - expx >= expk - MPFR_PREC(y) - 1)
    return 0;
  prec = (MPFR_PREC(y) + 2 <= ULSIZE) ? ULSIZE : MPFR_PREC(y) + 2;
  mpfr_init2 (t, prec);
  mpfr_set_ui (t, k, MPFR_RNDZ); /* exact since prec >= ULSIZE */
  mpfr_nextabove (t);
  /* now k < t < k + 2/x and round(t) = round(log2(1+x)) */
  inex = mpfr_set (y, t, rnd_mode);
  mpfr_clear (t);
  /* Warning: for RNDF, the mpfr_set calls above might return 0 */
  return (rnd_mode == MPFR_RNDF) ? 1 : inex;
}

/* The computation of log2p1 is done by log2p1(x) = log1p(x)/log(2) */
int
mpfr_log2p1 (mpfr_ptr y, mpfr_srcptr x, mpfr_rnd_t rnd_mode)
{
  int comp, inexact, nloop;
  mpfr_t t, lg2;
  mpfr_prec_t Ny = MPFR_PREC(y), prec;
  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_UNLIKELY (MPFR_IS_SINGULAR (x)))
    return mpfr_log1p (y, x, rnd_mode); /* same result for singular cases */

  comp = mpfr_cmp_si (x, -1);
  /* log2p1(x) is undefined for x < -1 */
  if (MPFR_UNLIKELY(comp <= 0))
    {
      if (comp == 0)
        /* x=0: log2p1(-1)=-inf (divide-by-zero exception) */
        {
          MPFR_SET_INF (y);
          MPFR_SET_NEG (y);
          MPFR_SET_DIVBY0 ();
          MPFR_RET (0);
        }
      MPFR_SET_NAN (y);
      MPFR_RET_NAN;
    }

  MPFR_SAVE_EXPO_MARK (expo);

  prec = Ny + MPFR_INT_CEIL_LOG2 (Ny) + 6;

  mpfr_init2 (t, prec);
  mpfr_init2 (lg2, prec);

  MPFR_ZIV_INIT (loop, prec);
  for (nloop = 0; ; nloop++)
    {
      mpfr_log1p (t, x, MPFR_RNDN);
      mpfr_const_log2 (lg2, MPFR_RNDN);
      mpfr_div (t, t, lg2, MPFR_RNDN);
      /* t = log2(1+x) * (1 + theta)^3 where |theta| < 2^-prec,
         for prec >= 2 we have |(1 + theta)^3 - 1| < 4*theta.
         Note: contrary to log10p1, no underflow is possible in extended
         exponent range, since for tiny x, |log2(1+x)| ~ |x|/log(2) >= |x|,
         and x is representable, thus x/log(2) too. */
      if (MPFR_LIKELY (MPFR_CAN_ROUND (t, prec - 2, Ny, rnd_mode)))
        break;

      if (nloop == 0)
        {
          /* check for exact cases */
          mpfr_exp_t k;

          MPFR_LOG_MSG (("check for exact cases\n", 0));
          k = mpfr_log2p1_isexact (x);
          if (k != 0) /* 1+x = 2^k */
            {
              inexact = mpfr_set_si (y, k, rnd_mode);
              goto end;
            }

          /* if x = 2^k with huge k, Ziv's loop will fail */
          inexact = mpfr_log2p1_special (y, x, rnd_mode);
          if (inexact != 0)
            goto end;
        }

      MPFR_ZIV_NEXT (loop, prec);
      mpfr_set_prec (t, prec);
      mpfr_set_prec (lg2, prec);
    }
  inexact = mpfr_set (y, t, rnd_mode);

 end:
  MPFR_ZIV_FREE (loop);
  mpfr_clear (t);
  mpfr_clear (lg2);

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