float.h
Go Up to C Run-Time Library Reference
float.h provides low-level functions to work with floating-point variables:
Functions
- _chgsign, _chgsignl
- _clear87, _clearfp
- _control87, _controlfp
- _copysign, _copysignl
- _finite, _finitel
- _fpclass, _fpclassl
- _fpreset
- _isnan, _isnanl
- _logb, _logbl
- _nextafter, _nextafterl
- _scalb, _scalbl
- _status87, _statusfp
Constants
Category | Constant | Value | Significance |
---|---|---|---|
FPU | CW_DEFAULT | _default87 | Default FPU control word: affine infinity, underflow and precision exceptions masked |
FLT_RADIX | 2 | ||
FLT_ROUNDS | 1 | ||
FLT_GUARD | 1 | ||
FLT_NORMALIZE | 1 | ||
Significant decimal digits | DBL_DIG | 15 | |
FLT_DIG | 6 | ||
LDBL_DIG | 18 | ||
Mantissa digits | DBL_MANT_DIG | 53 | Double precision mantissa digits |
FLT_MANT_DIG | 24 | Simple precision mantissa digits | |
LDBL_MANT_DIG | 64 | Extended precision mantissa digits | |
Machine epsilon | DBL_EPSILON | 2.2204460492503131E-16 | Double precision epsilon |
FLT_EPSILON | 1.19209290E-07F | Simple precision epsilon | |
LDBL_EPSILON | 1.084202172485504434e-019L | Extended precision epsilon | |
Smallest positive IEEE normal numbers | DBL_MIN | 2.2250738585072014E-308 | Smallest positive double precision number |
FLT_MIN | 1.17549435E-38F | Smallest positive simple precision number |