Each of the macros
isgreater(),
isgreaterequal(),
isless(),
islessequal(),
and
islessgreater()
take arguments
x
and
y
and return a non-zero value if and only if its nominal
relation on
x
and
y
is true.
These macros always return zero if either
argument is not a number (NaN), but unlike the corresponding C
operators, they never raise a floating point exception.
The
isunordered()
macro takes arguments
x
and
y
and returns non-zero if and only if any of
x
or
y
are NaNs.
For any pair of floating-point values, one
of the relationships (less, greater, equal, unordered) holds.
