Main Page | Namespace List | Class Hierarchy | Class List | Directories | File List | Namespace Members | Class Members | File Members | Search for 
Kernel

fix24.cpp File Reference

(r1785/r1282)

#include "camtypes.h"

Go to the source code of this file.

Functions

fixed24 Fixed24Mul (fixed24 a, fixed24 b)
 signed multiplication
fixed24 Fixed24Div (fixed24 a, fixed24 b)
 signed division
double Fixed24ToDouble (fixed24 a)
 General type conversion.
fixed24 DoubleToFixed24 (double arg)
 General type conversion.

Function Documentation

fixed24 DoubleToFixed24 (  double  arg  ) 
 

General type conversion.

Author:
Gavin_Theobald (Xara Group Ltd) <camelotdev@xara.com>
Date:
15/6/2005
Parameters:
double-precision argument [INPUTS]
None [OUTPUTS]
Returns:
FIXED24 value

Errors: None (overflow will cause exception)

Definition at line 207 of file fix24.cpp.

00208 {
00209     fixed24 result ;
00210 #if 0 //defined(_M_IX86)
00211     //
00212     // This avoids the slow fix operation. Add a large number so that
00213     // the double always has the same exponent and the mantissa becomes,
00214     // in affect, a fixed point value with the value we want in the
00215     // bottom 32 bits.
00216     //
00217     // Note that we are relying on the compiler to convert this to
00218     // efficient code.
00219     //
00220     // TODO: Test!
00221     //
00222     static const double fFix = (XLONG)3<<(DBL_MANT_DIG-24-2);
00223     double F = arg+fFix;
00224     result.all = (INT32&)F ;
00225 #else
00226     result.all = INT32(arg*(1<<24)) ;
00227 #endif
00228     return result ;
00229 }

fixed24 Fixed24Div (  fixed24  a,
fixed24  b
) 
 

signed division

Author:
Gavin_Theobald (Xara Group Ltd) <camelotdev@xara.com>
Date:
15/6/2005
Parameters:
two fixed24s [INPUTS]
None [OUTPUTS]
Returns:
FIXED24 result (arg1 / arg2) Scope: Only to be used in the FIXED24 class.

Errors: None (no overflow checks). No rounding either (should it?) Overflow will cause a DIV0 exception, as will divide by zero

Note that this is not very efficient as it forces a multiply that could instead be performed by a shift. The alternative is to use int64s but this would then compile as a 64bit/64bit division which is worse.

Why isn't there a Int64div32to32 function?

Definition at line 163 of file fix24.cpp.

00164 {
00165     fixed24 result ;
00166     result.all = MulDiv(a.all,1<<24,b.all) ;
00167     return result ;
00168 }

fixed24 Fixed24Mul (  fixed24  a,
fixed24  b
) 
 

signed multiplication

Author:
Gavin_Theobald (Xara Group Ltd) <camelotdev@xara.com>
Date:
15/6/2005
Parameters:
two fixed24s [INPUTS]
None [OUTPUTS]
Returns:
FIXED24 result (arg1 * arg2) Scope: Only to be used in the FIXED24 class.

Errors: None (no overflow checks). Rounds

Definition at line 133 of file fix24.cpp.

00134 {
00135     fixed24             result;
00136     result.all = INT32((Int32x32To64(a.all,b.all)+0x800000)>>24) ;
00137     return result;
00138 }

double Fixed24ToDouble (  fixed24  a  ) 
 

General type conversion.

Author:
Gavin_Theobald (Xara Group Ltd) <camelotdev@xara.com>
Date:
15/6/2005
Parameters:
FIXED24 argument [INPUTS]
None [OUTPUTS]
Returns:
double-precision value Scope: Only to be used in the FIXED24 class.

Errors: None

This code relies on the compiler efficiently replacing the division by a multiplication of the reciprical.

Definition at line 188 of file fix24.cpp.

00189 {
00190     return (double)a.all/(1<<24) ;
00191 }

Generated on Sat Nov 10 03:49:09 2007 for Camelot by  doxygen 1.4.4