Gets or sets the current floating-point rounding mode.

Syntax

int fegetround(void);int fesetround(int round_mode);

Parameters

round_mode
The rounding mode to set, as one of the floating-point rounding macros. If the value isn't equal to one of the floating-point rounding macros, the rounding mode isn't changed.

Return value

On success,fegetround returns the rounding mode as one of the floating point rounding macro values. It returns a negative value if the current rounding mode can't be determined.

On success,fesetround returns 0. Otherwise, a non-zero value is returned.

Remarks

Floating-point operations can use one of several rounding modes. These modes control which direction the results of floating-point operations are rounded toward when the results are stored. Here are the names and behaviors of the floating-point rounding macros defined in <fenv.h>:

The default behavior ofFE_TONEAREST is to round results midway between representable values toward the nearest value with an even (0) least significant bit.

The current rounding mode affects these operations:

• String conversions.
• The results of floating-point arithmetic operators outside of constant expressions.
• The library rounding functions, such asrint andnearbyint.
• Return values from standard library mathematical functions.

The current rounding mode doesn't affect these operations:

• Thetrunc,ceil,floor, andlround library functions.
• Floating-point to integer implicit casts and conversions, which always round towards zero.
• The results of floating-point arithmetic operators in constant expressions, which always round to the nearest value.

To use these functions, you must turn off floating-point optimizations that could prevent access by using the#pragma fenv_access(on) directive prior to the call. For more information, seefenv_access.

Requirements

For more information, seeCompatibility.

See also

Alphabetical function reference
nearbyint,nearbyintf,nearbyintl
rint,rintf,rintl
lrint,lrintf,lrintl,llrint,llrintf,llrintl