man3/nextafter.3

   1 .\" Copyright 2002 Walter Harms (walter.harms@informatik.uni-oldenburg.de)
   2 .\" Distributed under GPL
   3 .\" Based on glibc infopages
   4 .TH NEXTAFTER 3 2002-08-10 "GNU" "libc math functions"
   5 .SH NAME
   6 nextafter, nextafterf, nextafterl, nexttoward, nexttowardf, nexttowardl \- floating point number manipulation
   7 .SH SYNOPSIS
   8 .B #include <math.h>
   9 .sp
  10 .BI "double nextafter(double " x ", double " y );
  11 .br
  12 .BI "float nextafterf(float " x ", float " y );
  13 .br
  14 .BI "long double nextafterl(long double " x ", long double " y );
  15 .sp
  16 .BI "double nexttoward(double " x ", long double " y );
  17 .br
  18 .BI "float nexttowardf(float " x ", long double " y );
  19 .br
  20 .BI "long double nexttowardl(long double " x ", long double " y );
  21 .sp
  22 Link with \-lm.
  23 .SH DESCRIPTION
  24 The
  25 .B nextafter()
  26 functions return the next representable neighbor of
  27 \fIx\fP in the direction towards \fIy\fP.  The size of the step
  28 between \fIx\fP and the result depends on the type of the result.
  29 If \fIx\fP = \fIy\fP the function simply returns \fIy\fP.
  30 If either value is
  31 .IR NaN ,
  32 then
  33 .I NaN
  34 is returned.  Otherwise a value corresponding to the value of the
  35 least significant bit in the mantissa is added or subtracted,
  36 depending on the direction.
  37 .PP
  38 The
  39 .B nexttoward()
  40 functions do the same as the
  41 .B nextafter()
  42 functions, except that they have a long double second argument.
  43 .PP
  44 These functions will signal overflow or underflow if the result
  45 goes outside of the range of normalized numbers.
  46 .SH "CONFORMING TO"
  47 C99. This function is defined in IEC 559 (and the appendix with
  48 recommended functions in IEEE 754/IEEE 854).
  49 .SH "SEE ALSO"
  50 .BR nearbyint (3)