fxfixed.h File Reference

#include "brtypes.h"

Include dependency graph for fxfixed.h:

This graph shows which files directly or indirectly include this file:

Classes
class	Burger::IntMath
	Helper routines for integer math. More...
class	Burger::FixedMath
	Functions to handle fixed point integer math. More...
class	Burger::FracMath
	Functions to handle Frac point integer math. More...
struct	Burger::FixedVector2D_t
	A 2D Fixed32 vector. More...
struct	Burger::FixedVector3D_t
	3D Fixed32 vector. More...
struct	Burger::FixedVector4D_t
	4D Fixed32 vector. More...
struct	Burger::FixedMatrix3D_t
	3x3 Fixed32 matrix. More...
struct	Burger::FixedMatrix4D_t
	4x4 Fixed32 matrix. More...
Namespaces
namespace	Burger
Defines
#define	FLOATTOFIXED(x)   static_cast<Fixed32>((x)*65536.0f)
	Define to convert a float into a Fixed32.
#define	FIXEDTOFLOAT(x)   static_cast<float>((x)*(1.0f/65536.0f))
	Define to convert a Fixed32 into a float.
#define	INTTOFIXED(x)   static_cast<Fixed32>((x)<<16)
	Define to convert a integer into a Fixed32.
#define	FIXEDTOINT(x)   static_cast<int>((x+((x>>31)&0xFFFF))>>16)
	Define to convert a Fixed32 value to an integer.
Typedefs
typedef char	Burger::Float80Bit [10]
	80 bit float (Extended)
Variables
const float	g_fBurgerIntMathNearesttable [2]
	Helper constants for float to integer conversions.
const float	g_fBurgerMath65536
	Floating point constant of 65536.0f.
const Word16	Burger::MaxWord16 = 0xFFFFU
	Maximum value for Word16 (Minimum is 0)
const Word32	Burger::MaxWord32 = 0xFFFFFFFFUL
	Maximum value for Word32 (Minimum is 0)
const Int16	Burger::MinInt16 = (-0x7FFF)-1
	Minimum value for Int16.
const Int16	Burger::MaxInt16 = 0x7FFF
	Maximum value for Int16.
const Int32	Burger::MinInt32 = (-0x7FFFFFFFL)-1
	Minimum value for Int32.
const Int32	Burger::MaxInt32 = 0x7FFFFFFFL
	Maximum value for Int32.
const Frac32	Burger::MinFrac32 = (-0x7FFFFFFFL)-1
	Minimum value for Frac32.
const Frac32	Burger::MaxFrac32 = 0x7FFFFFFFL
	Maximum value for Frac32.
const Fixed32	Burger::MinFixed32 = (-0x7FFFFFFFL)-1
	Minimum value for Fixed32.
const Fixed32	Burger::MaxFixed32 = 0x7FFFFFFFL
	Maximum value for Fixed32.
const Fixed32	Burger::PiFixed32 = 0x0032440L
	Pi in Fixed32 format (3.141...)

---

Define Documentation

#define FIXEDTOFLOAT

(

x

)

static_cast<float>((x)*(1.0f/65536.0f))

Define to convert a Fixed32 into a float.

Macro to convert at compile time a 16.16 fixed point number into a floating point value Used for compile assignment of float constants using Fixed32 constants..

Parameters:

x	16.16 fixed point value to convert to a float.

Note:

Due to the integer to float conversion, use of this macro at runtime may be a performance penalty on the Xbox 360 and PS3.

    static const float MyData = FIXEDTOFLOAT(Burger::PiFixed32);

#define FIXEDTOINT

(

x

)

static_cast<int>((x+((x>>31)&0xFFFF))>>16)

Define to convert a Fixed32 value to an integer.

Macro to convert at compile time a 16.16 fixed point value into an integer. Used for compile time assignment of integers using Fixed32 constants.

Parameters:

x	Fixed32 value to convert to an integer.

Note:

This macro uses round to zero in the removal of the fraction. Negative numbers like -0.8 become 0, as well as 0.8 becomes 0.

    static const int MyData = FIXEDTOINT(Burger::PiFixed32);

#define FLOATTOFIXED

(

x

)

static_cast<Fixed32>((x)*65536.0f)

Define to convert a float into a Fixed32.

Macro to convert at compile time a floating point value into 16.16 fixed point. Used for compile assignment of Fixed32 constants.

Parameters:

x	Floating point value to convert to a Fixed32.

Note:

Due to the float to integer conversion, use of this macro at runtime may be a performance penalty on the Xbox 360 and PS3.

    static const Fixed32 MyData = FLOATTOFIXED(4.503f);

#define INTTOFIXED

(

x

)

static_cast<Fixed32>((x)<<16)

Define to convert a integer into a Fixed32.

Macro to convert at compile time an integer value into 16.16 fixed point. Used for compile assignment of Fixed32 constants.

Parameters:

x	Integer value to convert to a Fixed32.

Note:

There is no bounds checking performed by this macro. As a result, integers larger than 32767 or less than -32768 will call undefined behavior due to over/underflow.

    static const Fixed32 MyData = INTTOFIXED(443);

---

Variable Documentation

const float g_fBurgerIntMathNearesttable

Helper constants for float to integer conversions.

These constants are used for the translation of floats to ints using floor, ceil or zero. The first constant is added to yield floor() and the second is for getting ceil() (1.0f/(65536.0f*256.0f)) is the 32-bit floating point epsilon just under 1.0f

See also:

Burger::IntMath::FromFloat(float) or Burger::IntMath::FromFloatFloor(float)

const float g_fBurgerMath65536

Floating point constant of 65536.0f.