man3/ilogb.3

   1 .\" Copyright 2004 Andries Brouwer <aeb@cwi.nl>.
   2 .\" and Copyright 2008, Linux Foundation, written by Michael Kerrisk
   3 .\"     <mtk.manpages@gmail.com>
   4 .\"
   5 .\" SPDX-License-Identifier: Linux-man-pages-copyleft
   6 .\"
   7 .\" Inspired by a page by Walter Harms created 2002-08-10
   8 .\"
   9 .TH ILOGB 3 2021-03-22 "Linux man-pages (unreleased)"
  10 .SH NAME
  11 ilogb, ilogbf, ilogbl \- get integer exponent of a floating-point value
  12 .SH LIBRARY
  13 Math library
  14 .RI ( libm ", " \-lm )
  15 .SH SYNOPSIS
  16 .nf
  17 .B #include <math.h>
  18 .PP
  19 .BI "int ilogb(double " x );
  20 .BI "int ilogbf(float " x );
  21 .BI "int ilogbl(long double " x );
  22 .fi
  23 .PP
  24 .RS -4
  25 Feature Test Macro Requirements for glibc (see
  26 .BR feature_test_macros (7)):
  27 .RE
  28 .PP
  29 .BR ilogb ():
  30 .nf
  31     _ISOC99_SOURCE || _POSIX_C_SOURCE >= 200112L
  32         || _XOPEN_SOURCE >= 500
  33 .\"    || _XOPEN_SOURCE && _XOPEN_SOURCE_EXTENDED
  34         || /* Since glibc 2.19: */ _DEFAULT_SOURCE
  35         || /* Glibc <= 2.19: */ _BSD_SOURCE || _SVID_SOURCE
  36 .fi
  37 .PP
  38 .BR ilogbf (),
  39 .BR ilogbl ():
  40 .nf
  41     _ISOC99_SOURCE || _POSIX_C_SOURCE >= 200112L
  42         || /* Since glibc 2.19: */ _DEFAULT_SOURCE
  43         || /* Glibc <= 2.19: */ _BSD_SOURCE || _SVID_SOURCE
  44 .fi
  45 .SH DESCRIPTION
  46 These functions return the exponent part of their argument
  47 as a signed integer.
  48 When no error occurs, these functions
  49 are equivalent to the corresponding
  50 .BR logb (3)
  51 functions, cast to
  52 .IR int .
  53 .SH RETURN VALUE
  54 On success, these functions return the exponent of
  55 .IR x ,
  56 as a signed integer.
  57 .PP
  58 If
  59 .I x
  60 is zero, then a domain error occurs, and the functions return
  61 .\" the POSIX.1 spec for logb() says logb() gives pole error for this
  62 .\" case, but for ilogb() it says domain error.
  63 .BR FP_ILOGB0 .
  64 .\" glibc: The numeric value is either `INT_MIN' or `-INT_MAX'.
  65 .PP
  66 If
  67 .I x
  68 is a NaN, then a domain error occurs, and the functions return
  69 .BR FP_ILOGBNAN .
  70 .\" glibc: The numeric value is either `INT_MIN' or `INT_MAX'.
  71 .\" On i386, FP_ILOGB0 and FP_ILOGBNAN have the same value.
  72 .PP
  73 If
  74 .I x
  75 is negative infinity or positive infinity, then
  76 a domain error occurs, and the functions return
  77 .BR INT_MAX .
  78 .\"
  79 .\" POSIX.1-2001 also says:
  80 .\" If the correct value is greater than {INT_MAX}, {INT_MAX}
  81 .\" shall be returned and a domain error shall occur.
  82 .\"
  83 .\" If the correct value is less than {INT_MIN}, {INT_MIN}
  84 .\" shall be returned and a domain error shall occur.
  85 .SH ERRORS
  86 See
  87 .BR math_error (7)
  88 for information on how to determine whether an error has occurred
  89 when calling these functions.
  90 .PP
  91 The following errors can occur:
  92 .TP
  93 Domain error: \fIx\fP is 0 or a NaN
  94 An invalid floating-point exception
  95 .RB ( FE_INVALID )
  96 is raised, and
  97 .I errno
  98 is set to
  99 .B EDOM
 100 (but see BUGS).
 101 .TP
 102 Domain error: \fIx\fP is an infinity
 103 An invalid floating-point exception
 104 .RB ( FE_INVALID )
 105 is raised, and
 106 .I errno
 107 is set to
 108 .B EDOM
 109 (but see BUGS).
 110 .SH ATTRIBUTES
 111 For an explanation of the terms used in this section, see
 112 .BR attributes (7).
 113 .ad l
 114 .nh
 115 .TS
 116 allbox;
 117 lbx lb lb
 118 l l l.
 119 Interface       Attribute       Value
 120 T{
 121 .BR ilogb (),
 122 .BR ilogbf (),
 123 .BR ilogbl ()
 124 T}      Thread safety   MT-Safe
 125 .TE
 126 .hy
 127 .ad
 128 .sp 1
 129 .SH STANDARDS
 130 C99, POSIX.1-2001, POSIX.1-2008.
 131 .SH BUGS
 132 .\" Bug raised: http://sources.redhat.com/bugzilla/show_bug.cgi?id=6794
 133 Before version 2.16, the following bugs existed in the
 134 glibc implementation of these functions:
 135 .IP * 3
 136 The domain error case where
 137 .I x
 138 is 0 or a NaN did not cause
 139 .I errno
 140 to be set or (on some architectures) raise a floating-point exception.
 141 .IP * 3
 142 The domain error case where
 143 .I x
 144 is an infinity did not cause
 145 .I errno
 146 to be set or raise a floating-point exception.
 147 .SH SEE ALSO
 148 .BR log (3),
 149 .BR logb (3),
 150 .BR significand (3)