NAME

       fmax, fmaxf, fmaxl - determine maximum of two floating-point numbers

SYNOPSIS

       #include <math.h>

       double  fmax(double  x,  double  y); float fmaxf(float x, float y); long double fmaxl(long
       double x, long double y);

       Link with -lm.

   Feature Test Macro Requirements for glibc (see feature_test_macros(7)):

       fmax(), fmaxf(), fmaxl():
           _ISOC99_SOURCE || _POSIX_C_SOURCE >= 200112L

DESCRIPTION

       These functions return the larger value of x and y.

RETURN VALUE

       These functions return the maximum of x and y.

       If one argument is a NaN, the other argument is returned.

       If both arguments are NaN, a NaN is returned.

ERRORS

       No errors occur.

VERSIONS

       These functions first appeared in glibc in version 2.1.

ATTRIBUTES

       For an explanation of the terms used in this section, see attributes(7).

       ┌─────────────────────────┬───────────────┬─────────┐
       │Interface                │ Attribute     │ Value   │
       ├─────────────────────────┼───────────────┼─────────┤
       │fmax(), fmaxf(), fmaxl() │ Thread safety │ MT-Safe │
       └─────────────────────────┴───────────────┴─────────┘

CONFORMING TO

       C99, POSIX.1-2001, POSIX.1-2008.

SEE ALSO

       fdim(3), fmin(3)

COLOPHON

       This page is part of release 4.15 of the Linux man-pages project.  A  description  of  the
       project,  information  about  reporting  bugs, and the latest version of this page, can be
       found at https://www.kernel.org/doc/man-pages/.

                                            2017-09-15                                    FMAX(3)