freerdp/winpr/libwinpr/utils/trio/trionan.c

1169 lines
29 KiB
C
Raw Normal View History

2023-05-09 21:29:50 +00:00
/*************************************************************************
*
* $Id: trionan.c,v 1.33 2005/05/29 11:57:25 breese Exp $
*
* Copyright (C) 2001 Bjorn Reese <breese@users.sourceforge.net>
*
* Permission to use, copy, modify, and distribute this software for any
* purpose with or without fee is hereby granted, provided that the above
* copyright notice and this permission notice appear in all copies.
*
* THIS SOFTWARE IS PROVIDED ``AS IS'' AND WITHOUT ANY EXPRESS OR IMPLIED
* WARRANTIES, INCLUDING, WITHOUT LIMITATION, THE IMPLIED WARRANTIES OF
* MERCHANTIBILITY AND FITNESS FOR A PARTICULAR PURPOSE. THE AUTHORS AND
* CONTRIBUTORS ACCEPT NO RESPONSIBILITY IN ANY CONCEIVABLE MANNER.
*
************************************************************************
*
* Functions to handle special quantities in floating-point numbers
* (that is, NaNs and infinity). They provide the capability to detect
* and fabricate special quantities.
*
* Although written to be as portable as possible, it can never be
* guaranteed to work on all platforms, as not all hardware supports
* special quantities.
*
* The approach used here (approximately) is to:
*
* 1. Use C99 functionality when available.
* 2. Use IEEE 754 bit-patterns if possible.
* 3. Use platform-specific techniques.
*
************************************************************************/
/*************************************************************************
* Include files
*/
#include "triodef.h"
#include "trionan.h"
#include <math.h>
#include <string.h>
#include <limits.h>
#if !defined(TRIO_PLATFORM_SYMBIAN)
#include <float.h>
#endif
#if defined(TRIO_PLATFORM_UNIX)
#include <signal.h>
#endif
#if defined(TRIO_COMPILER_DECC)
#include <fp_class.h>
#endif
#include <assert.h>
#if defined(TRIO_DOCUMENTATION)
#include "doc/doc_nan.h"
#endif
/** @addtogroup SpecialQuantities
@{
*/
/*************************************************************************
* Definitions
*/
#if !defined(TRIO_PUBLIC_NAN)
#define TRIO_PUBLIC_NAN TRIO_PUBLIC
#endif
#if !defined(TRIO_PRIVATE_NAN)
#define TRIO_PRIVATE_NAN TRIO_PRIVATE
#endif
#define TRIO_TRUE (1 == 1)
#define TRIO_FALSE (0 == 1)
/*
* We must enable IEEE floating-point on Alpha
*/
#if defined(__alpha) && !defined(_IEEE_FP)
#if defined(TRIO_COMPILER_DECC)
#if defined(TRIO_PLATFORM_VMS)
#error "Must be compiled with option /IEEE_MODE=UNDERFLOW_TO_ZERO/FLOAT=IEEE"
#else
#if !defined(_CFE)
#error "Must be compiled with option -ieee"
#endif
#endif
#else
#if defined(TRIO_COMPILER_GCC)
#error "Must be compiled with option -mieee"
#endif
#endif
#endif /* __alpha && ! _IEEE_FP */
/*
* In ANSI/IEEE 754-1985 64-bits double format numbers have the
* following properties (amoungst others)
*
* o FLT_RADIX == 2: binary encoding
* o DBL_MAX_EXP == 1024: 11 bits exponent, where one bit is used
* to indicate special numbers (e.g. NaN and Infinity), so the
* maximum exponent is 10 bits wide (2^10 == 1024).
* o DBL_MANT_DIG == 53: The mantissa is 52 bits wide, but because
* numbers are normalized the initial binary 1 is represented
* implicitly (the so-called "hidden bit"), which leaves us with
* the ability to represent 53 bits wide mantissa.
*/
#if defined(__STDC_IEC_559__)
#define TRIO_IEEE_754
#else
#if (FLT_RADIX - 0 == 2) && (DBL_MAX_EXP - 0 == 1024) && (DBL_MANT_DIG - 0 == 53)
#define TRIO_IEEE_754
#endif
#endif
/*
* Determine which fpclassify_and_sign() function to use.
*/
#if defined(TRIO_FUNC_FPCLASSIFY_AND_SIGNBIT)
#if defined(PREDEF_STANDARD_C99) && defined(fpclassify)
#define TRIO_FUNC_C99_FPCLASSIFY_AND_SIGNBIT
#else
#if defined(TRIO_COMPILER_DECC)
#define TRIO_FUNC_DECC_FPCLASSIFY_AND_SIGNBIT
#else
#if defined(TRIO_COMPILER_VISUALC) || defined(TRIO_COMPILER_BORLAND)
#define TRIO_FUNC_MS_FPCLASSIFY_AND_SIGNBIT
#else
#if defined(TRIO_COMPILER_HP) && defined(FP_PLUS_NORM)
#define TRIO_FUNC_HP_FPCLASSIFY_AND_SIGNBIT
#else
#if defined(TRIO_COMPILER_XLC) && defined(FP_PLUS_NORM)
#define TRIO_FUNC_XLC_FPCLASSIFY_AND_SIGNBIT
#else
#define TRIO_FUNC_INTERNAL_FPCLASSIFY_AND_SIGNBIT
#endif
#endif
#endif
#endif
#endif
#endif
/*
* Determine how to generate negative zero.
*/
#if defined(TRIO_FUNC_NZERO)
#if defined(TRIO_IEEE_754)
#define TRIO_NZERO_IEEE_754
#else
#define TRIO_NZERO_FALLBACK
#endif
#endif
/*
* Determine how to generate positive infinity.
*/
#if defined(TRIO_FUNC_PINF)
#if defined(INFINITY) && defined(__STDC_IEC_559__)
#define TRIO_PINF_C99_MACRO
#else
#if defined(TRIO_IEEE_754)
#define TRIO_PINF_IEEE_754
#else
#define TRIO_PINF_FALLBACK
#endif
#endif
#endif
/*
* Determine how to generate NaN.
*/
#if defined(TRIO_FUNC_NAN)
#if defined(PREDEF_STANDARD_C99) && !defined(TRIO_COMPILER_DECC)
#define TRIO_NAN_C99_FUNCTION
#else
#if defined(NAN) && defined(__STDC_IEC_559__)
#define TRIO_NAN_C99_MACRO
#else
#if defined(TRIO_IEEE_754)
#define TRIO_NAN_IEEE_754
#else
#define TRIO_NAN_FALLBACK
#endif
#endif
#endif
#endif
/*
* Resolve internal dependencies.
*/
#if defined(TRIO_FUNC_INTERNAL_FPCLASSIFY_AND_SIGNBIT)
#define TRIO_FUNC_INTERNAL_ISNAN
#define TRIO_FUNC_INTERNAL_ISINF
#if defined(TRIO_IEEE_754)
#define TRIO_FUNC_INTERNAL_IS_SPECIAL_QUANTITY
#define TRIO_FUNC_INTERNAL_IS_NEGATIVE
#endif
#endif
#if defined(TRIO_NZERO_IEEE_754) || defined(TRIO_PINF_IEEE_754) || defined(TRIO_NAN_IEEE_754)
#define TRIO_FUNC_INTERNAL_MAKE_DOUBLE
#endif
#if defined(TRIO_FUNC_INTERNAL_ISNAN)
#if defined(PREDEF_STANDARD_XPG3)
#define TRIO_INTERNAL_ISNAN_XPG3
#else
#if defined(TRIO_IEEE_754)
#define TRIO_INTERNAL_ISNAN_IEEE_754
#else
#define TRIO_INTERNAL_ISNAN_FALLBACK
#endif
#endif
#endif
#if defined(TRIO_FUNC_INTERNAL_ISINF)
#if defined(TRIO_IEEE_754)
#define TRIO_INTERNAL_ISINF_IEEE_754
#else
#define TRIO_INTERNAL_ISINF_FALLBACK
#endif
#endif
/*************************************************************************
* Constants
*/
#if !defined(TRIO_EMBED_NAN)
/* Unused but kept for reference */
/* static TRIO_CONST char rcsid[] = "@(#)$Id: trionan.c,v 1.33 2005/05/29 11:57:25 breese Exp $"; */
#endif
#if defined(TRIO_FUNC_INTERNAL_MAKE_DOUBLE) || defined(TRIO_FUNC_INTERNAL_IS_SPECIAL_QUANTITY) || \
defined(TRIO_FUNC_INTERNAL_IS_NEGATIVE)
/*
* Endian-agnostic indexing macro.
*
* The value of internalEndianMagic, when converted into a 64-bit
* integer, becomes 0x0706050403020100 (we could have used a 64-bit
* integer value instead of a double, but not all platforms supports
* that type). The value is automatically encoded with the correct
* endianess by the compiler, which means that we can support any
* kind of endianess. The individual bytes are then used as an index
* for the IEEE 754 bit-patterns and masks.
*/
#define TRIO_DOUBLE_INDEX(x) (((unsigned char*)&internalEndianMagic)[7 - (x)])
static TRIO_CONST double internalEndianMagic = 7.949928895127363e-275;
#endif
#if defined(TRIO_FUNC_INTERNAL_IS_SPECIAL_QUANTITY)
/* Mask for the exponent */
static TRIO_CONST unsigned char ieee_754_exponent_mask[] = { 0x7F, 0xF0, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00 };
/* Mask for the mantissa */
static TRIO_CONST unsigned char ieee_754_mantissa_mask[] = { 0x00, 0x0F, 0xFF, 0xFF,
0xFF, 0xFF, 0xFF, 0xFF };
#endif
#if defined(TRIO_FUNC_INTERNAL_IS_NEGATIVE)
/* Mask for the sign bit */
static TRIO_CONST unsigned char ieee_754_sign_mask[] = { 0x80, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00 };
#endif
#if defined(TRIO_NZERO_IEEE_754)
/* Bit-pattern for negative zero */
static TRIO_CONST unsigned char ieee_754_negzero_array[] = { 0x80, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00 };
#endif
#if defined(TRIO_PINF_IEEE_754)
/* Bit-pattern for infinity */
static TRIO_CONST unsigned char ieee_754_infinity_array[] = { 0x7F, 0xF0, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00 };
#endif
#if defined(TRIO_NAN_IEEE_754)
/* Bit-pattern for quiet NaN */
static TRIO_CONST unsigned char ieee_754_qnan_array[] = { 0x7F, 0xF8, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00 };
#endif
/*************************************************************************
* Internal functions
*/
/*
* internal_make_double
*/
#if defined(TRIO_FUNC_INTERNAL_MAKE_DOUBLE)
TRIO_PRIVATE_NAN double internal_make_double TRIO_ARGS1((values), TRIO_CONST unsigned char* values)
{
TRIO_VOLATILE double result;
int i;
for (i = 0; i < (int)sizeof(double); i++)
{
((TRIO_VOLATILE unsigned char*)&result)[TRIO_DOUBLE_INDEX(i)] = values[i];
}
return result;
}
#endif
/*
* internal_is_special_quantity
*/
#if defined(TRIO_FUNC_INTERNAL_IS_SPECIAL_QUANTITY)
TRIO_PRIVATE_NAN int internal_is_special_quantity TRIO_ARGS2((number, has_mantissa), double number,
int* has_mantissa)
{
unsigned int i;
unsigned char current;
int is_special_quantity = TRIO_TRUE;
*has_mantissa = 0;
for (i = 0; i < (unsigned int)sizeof(double); i++)
{
current = ((unsigned char*)&number)[TRIO_DOUBLE_INDEX(i)];
is_special_quantity &= ((current & ieee_754_exponent_mask[i]) == ieee_754_exponent_mask[i]);
*has_mantissa |= (current & ieee_754_mantissa_mask[i]);
}
return is_special_quantity;
}
#endif
/*
* internal_is_negative
*/
#if defined(TRIO_FUNC_INTERNAL_IS_NEGATIVE)
TRIO_PRIVATE_NAN int internal_is_negative TRIO_ARGS1((number), double number)
{
unsigned int i;
int is_negative = TRIO_FALSE;
for (i = 0; i < (unsigned int)sizeof(double); i++)
{
is_negative |= (((unsigned char*)&number)[TRIO_DOUBLE_INDEX(i)] & ieee_754_sign_mask[i]);
}
return is_negative;
}
#endif
#if defined(TRIO_FUNC_C99_FPCLASSIFY_AND_SIGNBIT)
TRIO_PRIVATE_NAN TRIO_INLINE int c99_fpclassify_and_signbit TRIO_ARGS2((number, is_negative),
double number,
int* is_negative)
{
*is_negative = signbit(number);
switch (fpclassify(number))
{
case FP_NAN:
return TRIO_FP_NAN;
case FP_INFINITE:
return TRIO_FP_INFINITE;
case FP_SUBNORMAL:
return TRIO_FP_SUBNORMAL;
case FP_ZERO:
return TRIO_FP_ZERO;
default:
return TRIO_FP_NORMAL;
}
}
#endif /* TRIO_FUNC_C99_FPCLASSIFY_AND_SIGNBIT */
#if defined(TRIO_FUNC_DECC_FPCLASSIFY_AND_SIGNBIT)
TRIO_PRIVATE_NAN TRIO_INLINE int decc_fpclassify_and_signbit TRIO_ARGS2((number, is_negative),
double number,
int* is_negative)
{
switch (fp_class(number))
{
case FP_QNAN:
case FP_SNAN:
*is_negative = TRIO_FALSE; /* NaN has no sign */
return TRIO_FP_NAN;
case FP_POS_INF:
*is_negative = TRIO_FALSE;
return TRIO_FP_INFINITE;
case FP_NEG_INF:
*is_negative = TRIO_TRUE;
return TRIO_FP_INFINITE;
case FP_POS_DENORM:
*is_negative = TRIO_FALSE;
return TRIO_FP_SUBNORMAL;
case FP_NEG_DENORM:
*is_negative = TRIO_TRUE;
return TRIO_FP_SUBNORMAL;
case FP_POS_ZERO:
*is_negative = TRIO_FALSE;
return TRIO_FP_ZERO;
case FP_NEG_ZERO:
*is_negative = TRIO_TRUE;
return TRIO_FP_ZERO;
case FP_POS_NORM:
*is_negative = TRIO_FALSE;
return TRIO_FP_NORMAL;
case FP_NEG_NORM:
*is_negative = TRIO_TRUE;
return TRIO_FP_NORMAL;
default:
*is_negative = (number < 0.0);
return TRIO_FP_NORMAL;
}
}
#endif /* TRIO_FUNC_DECC_FPCLASSIFY_AND_SIGNBIT */
#if defined(TRIO_FUNC_MS_FPCLASSIFY_AND_SIGNBIT)
TRIO_PRIVATE_NAN int ms_fpclassify_and_signbit TRIO_ARGS2((number, is_negative), double number,
int* is_negative)
{
int result;
#if defined(TRIO_COMPILER_BORLAND)
/*
* The floating-point precision may be changed by the Borland _fpclass()
* function, so we have to save and restore the floating-point control mask.
*/
unsigned int mask;
/* Remember the old mask */
mask = _control87(0, 0);
#endif
switch (_fpclass(number))
{
case _FPCLASS_QNAN:
case _FPCLASS_SNAN:
*is_negative = TRIO_FALSE; /* NaN has no sign */
result = TRIO_FP_NAN;
break;
case _FPCLASS_PINF:
*is_negative = TRIO_FALSE;
result = TRIO_FP_INFINITE;
break;
case _FPCLASS_NINF:
*is_negative = TRIO_TRUE;
result = TRIO_FP_INFINITE;
break;
case _FPCLASS_PD:
*is_negative = TRIO_FALSE;
result = TRIO_FP_SUBNORMAL;
break;
case _FPCLASS_ND:
*is_negative = TRIO_TRUE;
result = TRIO_FP_SUBNORMAL;
break;
case _FPCLASS_PZ:
*is_negative = TRIO_FALSE;
result = TRIO_FP_ZERO;
break;
case _FPCLASS_NZ:
*is_negative = TRIO_TRUE;
result = TRIO_FP_ZERO;
break;
case _FPCLASS_PN:
*is_negative = TRIO_FALSE;
result = TRIO_FP_NORMAL;
break;
case _FPCLASS_NN:
*is_negative = TRIO_TRUE;
result = TRIO_FP_NORMAL;
break;
default:
*is_negative = (number < 0.0);
result = TRIO_FP_NORMAL;
break;
}
#if defined(TRIO_COMPILER_BORLAND)
/* Restore the old precision */
(void)_control87(mask, MCW_PC);
#endif
return result;
}
#endif /* TRIO_FUNC_MS_FPCLASSIFY_AND_SIGNBIT */
#if defined(TRIO_FUNC_HP_FPCLASSIFY_AND_SIGNBIT)
TRIO_PRIVATE_NAN TRIO_INLINE int hp_fpclassify_and_signbit TRIO_ARGS2((number, is_negative),
double number,
int* is_negative)
{
/*
* HP-UX 9.x and 10.x have an fpclassify() function, that is different
* from the C99 fpclassify() macro supported on HP-UX 11.x.
*/
switch (fpclassify(number))
{
case FP_QNAN:
case FP_SNAN:
*is_negative = TRIO_FALSE; /* NaN has no sign */
return TRIO_FP_NAN;
case FP_PLUS_INF:
*is_negative = TRIO_FALSE;
return TRIO_FP_INFINITE;
case FP_MINUS_INF:
*is_negative = TRIO_TRUE;
return TRIO_FP_INFINITE;
case FP_PLUS_DENORM:
*is_negative = TRIO_FALSE;
return TRIO_FP_SUBNORMAL;
case FP_MINUS_DENORM:
*is_negative = TRIO_TRUE;
return TRIO_FP_SUBNORMAL;
case FP_PLUS_ZERO:
*is_negative = TRIO_FALSE;
return TRIO_FP_ZERO;
case FP_MINUS_ZERO:
*is_negative = TRIO_TRUE;
return TRIO_FP_ZERO;
case FP_PLUS_NORM:
*is_negative = TRIO_FALSE;
return TRIO_FP_NORMAL;
case FP_MINUS_NORM:
*is_negative = TRIO_TRUE;
return TRIO_FP_NORMAL;
default:
*is_negative = (number < 0.0);
return TRIO_FP_NORMAL;
}
}
#endif /* TRIO_FUNC_HP_FPCLASSIFY_AND_SIGNBIT */
#if defined(TRIO_FUNC_XLC_FPCLASSIFY_AND_SIGNBIT)
TRIO_PRIVATE_NAN TRIO_INLINE int xlc_fpclassify_and_signbit TRIO_ARGS2((number, is_negative),
double number,
int* is_negative)
{
/*
* AIX has class() for C, and _class() for C++
*/
#if defined(__cplusplus)
#define AIX_CLASS(n) _class(n)
#else
#define AIX_CLASS(n) class(n)
#endif
switch (AIX_CLASS(number))
{
case FP_QNAN:
case FP_SNAN:
*is_negative = TRIO_FALSE; /* NaN has no sign */
return TRIO_FP_NAN;
case FP_PLUS_INF:
*is_negative = TRIO_FALSE;
return TRIO_FP_INFINITE;
case FP_MINUS_INF:
*is_negative = TRIO_TRUE;
return TRIO_FP_INFINITE;
case FP_PLUS_DENORM:
*is_negative = TRIO_FALSE;
return TRIO_FP_SUBNORMAL;
case FP_MINUS_DENORM:
*is_negative = TRIO_TRUE;
return TRIO_FP_SUBNORMAL;
case FP_PLUS_ZERO:
*is_negative = TRIO_FALSE;
return TRIO_FP_ZERO;
case FP_MINUS_ZERO:
*is_negative = TRIO_TRUE;
return TRIO_FP_ZERO;
case FP_PLUS_NORM:
*is_negative = TRIO_FALSE;
return TRIO_FP_NORMAL;
case FP_MINUS_NORM:
*is_negative = TRIO_TRUE;
return TRIO_FP_NORMAL;
default:
*is_negative = (number < 0.0);
return TRIO_FP_NORMAL;
}
}
#endif /* TRIO_FUNC_XLC_FPCLASSIFY_AND_SIGNBIT */
#if defined(TRIO_FUNC_INTERNAL_ISNAN)
TRIO_PRIVATE_NAN TRIO_INLINE int internal_isnan TRIO_ARGS1((number), double number)
{
#if defined(TRIO_INTERNAL_ISNAN_XPG3) || defined(TRIO_PLATFORM_SYMBIAN)
/*
* XPG3 defines isnan() as a function.
*/
return isnan(number);
#endif
#if defined(TRIO_INTERNAL_ISNAN_IEEE_754)
/*
* Examine IEEE 754 bit-pattern. A NaN must have a special exponent
* pattern, and a non-empty mantissa.
*/
int has_mantissa;
int is_special_quantity;
is_special_quantity = internal_is_special_quantity(number, &has_mantissa);
return (is_special_quantity && has_mantissa);
#endif
#if defined(TRIO_INTERNAL_ISNAN_FALLBACK)
/*
* Fallback solution
*/
int status;
double integral, fraction;
#if defined(TRIO_PLATFORM_UNIX)
void (*signal_handler)(int) = signal(SIGFPE, SIG_IGN);
#endif
status = (/*
* NaN is the only number which does not compare to itself
*/
((TRIO_VOLATILE double)number != (TRIO_VOLATILE double)number) ||
/*
* Fallback solution if NaN compares to NaN
*/
((number != 0.0) && (fraction = modf(number, &integral), integral == fraction)));
#if defined(TRIO_PLATFORM_UNIX)
signal(SIGFPE, signal_handler);
#endif
return status;
#endif
}
#endif /* TRIO_FUNC_INTERNAL_ISNAN */
#if defined(TRIO_FUNC_INTERNAL_ISINF)
TRIO_PRIVATE_NAN TRIO_INLINE int internal_isinf TRIO_ARGS1((number), double number)
{
#if defined(TRIO_PLATFORM_SYMBIAN)
return isinf(number);
#endif
#if defined(TRIO_INTERNAL_ISINF_IEEE_754)
/*
* Examine IEEE 754 bit-pattern. Infinity must have a special exponent
* pattern, and an empty mantissa.
*/
int has_mantissa;
int is_special_quantity;
is_special_quantity = internal_is_special_quantity(number, &has_mantissa);
return (is_special_quantity && !has_mantissa) ? ((number < 0.0) ? -1 : 1) : 0;
#endif
#if defined(TRIO_INTERNAL_ISINF_FALLBACK)
/*
* Fallback solution.
*/
int status;
#if defined(TRIO_PLATFORM_UNIX)
void (*signal_handler)(int) = signal(SIGFPE, SIG_IGN);
#endif
double infinity = trio_pinf();
status = ((number == infinity) ? 1 : ((number == -infinity) ? -1 : 0));
#if defined(TRIO_PLATFORM_UNIX)
signal(SIGFPE, signal_handler);
#endif
return status;
#endif
}
#endif /* TRIO_FUNC_INTERNAL_ISINF */
/*************************************************************************
* Public functions
*/
#if defined(TRIO_FUNC_FPCLASSIFY_AND_SIGNBIT)
TRIO_PUBLIC_NAN int trio_fpclassify_and_signbit TRIO_ARGS2((number, is_negative), double number,
int* is_negative)
{
/* The TRIO_FUNC_xxx_FPCLASSIFY_AND_SIGNBIT macros are mutually exclusive */
#if defined(TRIO_FUNC_C99_FPCLASSIFY_AND_SIGNBIT)
return c99_fpclassify_and_signbit(number, is_negative);
#endif
#if defined(TRIO_FUNC_DECC_FPCLASSIFY_AND_SIGNBIT)
return decc_fpclassify_and_signbit(number, is_negative);
#endif
#if defined(TRIO_FUNC_MS_FPCLASSIFY_AND_SIGNBIT)
return ms_fpclassify_and_signbit(number, is_negative);
#endif
#if defined(TRIO_FUNC_HP_FPCLASSIFY_AND_SIGNBIT)
return hp_fpclassify_and_signbit(number, is_negative);
#endif
#if defined(TRIO_FUNC_XLC_FPCLASSIFY_AND_SIGNBIT)
return xlc_fpclassify_and_signbit(number, is_negative);
#endif
#if defined(TRIO_FUNC_INTERNAL_FPCLASSIFY_AND_SIGNBIT)
/*
* Fallback solution.
*/
int rc;
if (number == 0.0)
{
/*
* In IEEE 754 the sign of zero is ignored in comparisons, so we
* have to handle this as a special case by examining the sign bit
* directly.
*/
#if defined(TRIO_IEEE_754)
*is_negative = internal_is_negative(number);
#else
*is_negative = TRIO_FALSE; /* FIXME */
#endif
return TRIO_FP_ZERO;
}
if (internal_isnan(number))
{
*is_negative = TRIO_FALSE;
return TRIO_FP_NAN;
}
rc = internal_isinf(number);
if (rc != 0)
{
*is_negative = (rc == -1);
return TRIO_FP_INFINITE;
}
if ((number > 0.0) && (number < DBL_MIN))
{
*is_negative = TRIO_FALSE;
return TRIO_FP_SUBNORMAL;
}
if ((number < 0.0) && (number > -DBL_MIN))
{
*is_negative = TRIO_TRUE;
return TRIO_FP_SUBNORMAL;
}
*is_negative = (number < 0.0);
return TRIO_FP_NORMAL;
#endif
}
#endif
/**
Check for NaN.
@param number An arbitrary floating-point number.
@return Boolean value indicating whether or not the number is a NaN.
*/
#if defined(TRIO_FUNC_ISNAN)
TRIO_PUBLIC_NAN int trio_isnan TRIO_ARGS1((number), double number)
{
int dummy;
return (trio_fpclassify_and_signbit(number, &dummy) == TRIO_FP_NAN);
}
#endif
/**
Check for infinity.
@param number An arbitrary floating-point number.
@return 1 if positive infinity, -1 if negative infinity, 0 otherwise.
*/
#if defined(TRIO_FUNC_ISINF)
TRIO_PUBLIC_NAN int trio_isinf TRIO_ARGS1((number), double number)
{
int is_negative;
if (trio_fpclassify_and_signbit(number, &is_negative) == TRIO_FP_INFINITE)
{
return (is_negative) ? -1 : 1;
}
else
{
return 0;
}
}
#endif
/**
Check for finity.
@param number An arbitrary floating-point number.
@return Boolean value indicating whether or not the number is a finite.
*/
#if defined(TRIO_FUNC_ISFINITE)
TRIO_PUBLIC_NAN int trio_isfinite TRIO_ARGS1((number), double number)
{
int dummy;
switch (trio_fpclassify_and_signbit(number, &dummy))
{
case TRIO_FP_INFINITE:
case TRIO_FP_NAN:
return 0;
default:
return 1;
}
}
#endif
/**
Examine the sign of a number.
@param number An arbitrary floating-point number.
@return Boolean value indicating whether or not the number has the
sign bit set (i.e. is negative).
*/
#if defined(TRIO_FUNC_SIGNBIT)
TRIO_PUBLIC_NAN int trio_signbit TRIO_ARGS1((number), double number)
{
int is_negative;
(void)trio_fpclassify_and_signbit(number, &is_negative);
return is_negative;
}
#endif
/**
Examine the class of a number.
@param number An arbitrary floating-point number.
@return Enumerable value indicating the class of @p number
*/
#if defined(TRIO_FUNC_FPCLASSIFY)
TRIO_PUBLIC_NAN int trio_fpclassify TRIO_ARGS1((number), double number)
{
int dummy;
return trio_fpclassify_and_signbit(number, &dummy);
}
#endif
/**
Generate negative zero.
@return Floating-point representation of negative zero.
*/
#if defined(TRIO_FUNC_NZERO)
TRIO_PUBLIC_NAN double trio_nzero(TRIO_NOARGS)
{
#if defined(TRIO_NZERO_IEEE_754)
return internal_make_double(ieee_754_negzero_array);
#endif
#if defined(TRIO_NZERO_FALLBACK)
TRIO_VOLATILE double zero = 0.0;
return -zero;
#endif
}
#endif
/**
Generate positive infinity.
@return Floating-point representation of positive infinity.
*/
#if defined(TRIO_FUNC_PINF)
TRIO_PUBLIC_NAN double trio_pinf(TRIO_NOARGS)
{
/* Cache the result */
static double pinf_value = 0.0;
if (pinf_value == 0.0)
{
#if defined(TRIO_PINF_C99_MACRO)
pinf_value = (double)INFINITY;
#endif
#if defined(TRIO_PINF_IEEE_754)
pinf_value = internal_make_double(ieee_754_infinity_array);
#endif
#if defined(TRIO_PINF_FALLBACK)
/*
* If HUGE_VAL is different from DBL_MAX, then HUGE_VAL is used
* as infinity. Otherwise we have to resort to an overflow
* operation to generate infinity.
*/
#if defined(TRIO_PLATFORM_UNIX)
void (*signal_handler)(int) = signal(SIGFPE, SIG_IGN);
#endif
pinf_value = HUGE_VAL;
if (HUGE_VAL == DBL_MAX)
{
/* Force overflow */
pinf_value += HUGE_VAL;
}
#if defined(TRIO_PLATFORM_UNIX)
signal(SIGFPE, signal_handler);
#endif
#endif
}
return pinf_value;
}
#endif
/**
Generate negative infinity.
@return Floating-point value of negative infinity.
*/
#if defined(TRIO_FUNC_NINF)
TRIO_PUBLIC_NAN double trio_ninf(TRIO_NOARGS)
{
static double ninf_value = 0.0;
if (ninf_value == 0.0)
{
/*
* Negative infinity is calculated by negating positive infinity,
* which can be done because it is legal to do calculations on
* infinity (for example, 1 / infinity == 0).
*/
ninf_value = -trio_pinf();
}
return ninf_value;
}
#endif
/**
Generate NaN.
@return Floating-point representation of NaN.
*/
#if defined(TRIO_FUNC_NAN)
TRIO_PUBLIC_NAN double trio_nan(TRIO_NOARGS)
{
/* Cache the result */
static double nan_value = 0.0;
if (nan_value == 0.0)
{
#if defined(TRIO_NAN_C99_FUNCTION) || defined(TRIO_PLATFORM_SYMBIAN)
nan_value = nan("");
#endif
#if defined(TRIO_NAN_C99_MACRO)
nan_value = (double)NAN;
#endif
#if defined(TRIO_NAN_IEEE_754)
nan_value = internal_make_double(ieee_754_qnan_array);
#endif
#if defined(TRIO_NAN_FALLBACK)
/*
* There are several ways to generate NaN. The one used here is
* to divide infinity by infinity. I would have preferred to add
* negative infinity to positive infinity, but that yields wrong
* result (infinity) on FreeBSD.
*
* This may fail if the hardware does not support NaN, or if
* the Invalid Operation floating-point exception is unmasked.
*/
#if defined(TRIO_PLATFORM_UNIX)
void (*signal_handler)(int) = signal(SIGFPE, SIG_IGN);
#endif
nan_value = trio_pinf() / trio_pinf();
#if defined(TRIO_PLATFORM_UNIX)
signal(SIGFPE, signal_handler);
#endif
#endif
}
return nan_value;
}
#endif
/** @} SpecialQuantities */
/*************************************************************************
* For test purposes.
*
* Add the following compiler option to include this test code.
*
* Unix : -DSTANDALONE
* VMS : /DEFINE=(STANDALONE)
*/
#if defined(STANDALONE)
#include <stdio.h>
static TRIO_CONST char* getClassification TRIO_ARGS1((type), int type)
{
switch (type)
{
case TRIO_FP_INFINITE:
return "FP_INFINITE";
case TRIO_FP_NAN:
return "FP_NAN";
case TRIO_FP_NORMAL:
return "FP_NORMAL";
case TRIO_FP_SUBNORMAL:
return "FP_SUBNORMAL";
case TRIO_FP_ZERO:
return "FP_ZERO";
default:
return "FP_UNKNOWN";
}
}
static void print_class TRIO_ARGS2((prefix, number), TRIO_CONST char* prefix, double number)
{
printf("%-6s: %s %-15s %g\n", prefix, trio_signbit(number) ? "-" : "+",
getClassification(trio_fpclassify(number)), number);
}
int main(TRIO_NOARGS)
{
double my_nan;
double my_pinf;
double my_ninf;
#if defined(TRIO_PLATFORM_UNIX)
void(*signal_handler) TRIO_PROTO((int));
#endif
my_nan = trio_nan();
my_pinf = trio_pinf();
my_ninf = trio_ninf();
print_class("Nan", my_nan);
print_class("PInf", my_pinf);
print_class("NInf", my_ninf);
print_class("PZero", 0.0);
print_class("NZero", -0.0);
print_class("PNorm", 1.0);
print_class("NNorm", -1.0);
print_class("PSub", 1.01e-307 - 1.00e-307);
print_class("NSub", 1.00e-307 - 1.01e-307);
printf("NaN : %4g 0x%02x%02x%02x%02x%02x%02x%02x%02x (%2d, %2d, %2d)\n", my_nan,
((unsigned char*)&my_nan)[0], ((unsigned char*)&my_nan)[1], ((unsigned char*)&my_nan)[2],
((unsigned char*)&my_nan)[3], ((unsigned char*)&my_nan)[4], ((unsigned char*)&my_nan)[5],
((unsigned char*)&my_nan)[6], ((unsigned char*)&my_nan)[7], trio_isnan(my_nan),
trio_isinf(my_nan), trio_isfinite(my_nan));
printf("PInf: %4g 0x%02x%02x%02x%02x%02x%02x%02x%02x (%2d, %2d, %2d)\n", my_pinf,
((unsigned char*)&my_pinf)[0], ((unsigned char*)&my_pinf)[1],
((unsigned char*)&my_pinf)[2], ((unsigned char*)&my_pinf)[3],
((unsigned char*)&my_pinf)[4], ((unsigned char*)&my_pinf)[5],
((unsigned char*)&my_pinf)[6], ((unsigned char*)&my_pinf)[7], trio_isnan(my_pinf),
trio_isinf(my_pinf), trio_isfinite(my_pinf));
printf("NInf: %4g 0x%02x%02x%02x%02x%02x%02x%02x%02x (%2d, %2d, %2d)\n", my_ninf,
((unsigned char*)&my_ninf)[0], ((unsigned char*)&my_ninf)[1],
((unsigned char*)&my_ninf)[2], ((unsigned char*)&my_ninf)[3],
((unsigned char*)&my_ninf)[4], ((unsigned char*)&my_ninf)[5],
((unsigned char*)&my_ninf)[6], ((unsigned char*)&my_ninf)[7], trio_isnan(my_ninf),
trio_isinf(my_ninf), trio_isfinite(my_ninf));
#if defined(TRIO_PLATFORM_UNIX)
signal_handler = signal(SIGFPE, SIG_IGN);
#endif
my_pinf = DBL_MAX + DBL_MAX;
my_ninf = -my_pinf;
my_nan = my_pinf / my_pinf;
#if defined(TRIO_PLATFORM_UNIX)
signal(SIGFPE, signal_handler);
#endif
printf("NaN : %4g 0x%02x%02x%02x%02x%02x%02x%02x%02x (%2d, %2d, %2d)\n", my_nan,
((unsigned char*)&my_nan)[0], ((unsigned char*)&my_nan)[1], ((unsigned char*)&my_nan)[2],
((unsigned char*)&my_nan)[3], ((unsigned char*)&my_nan)[4], ((unsigned char*)&my_nan)[5],
((unsigned char*)&my_nan)[6], ((unsigned char*)&my_nan)[7], trio_isnan(my_nan),
trio_isinf(my_nan), trio_isfinite(my_nan));
printf("PInf: %4g 0x%02x%02x%02x%02x%02x%02x%02x%02x (%2d, %2d, %2d)\n", my_pinf,
((unsigned char*)&my_pinf)[0], ((unsigned char*)&my_pinf)[1],
((unsigned char*)&my_pinf)[2], ((unsigned char*)&my_pinf)[3],
((unsigned char*)&my_pinf)[4], ((unsigned char*)&my_pinf)[5],
((unsigned char*)&my_pinf)[6], ((unsigned char*)&my_pinf)[7], trio_isnan(my_pinf),
trio_isinf(my_pinf), trio_isfinite(my_pinf));
printf("NInf: %4g 0x%02x%02x%02x%02x%02x%02x%02x%02x (%2d, %2d, %2d)\n", my_ninf,
((unsigned char*)&my_ninf)[0], ((unsigned char*)&my_ninf)[1],
((unsigned char*)&my_ninf)[2], ((unsigned char*)&my_ninf)[3],
((unsigned char*)&my_ninf)[4], ((unsigned char*)&my_ninf)[5],
((unsigned char*)&my_ninf)[6], ((unsigned char*)&my_ninf)[7], trio_isnan(my_ninf),
trio_isinf(my_ninf), trio_isfinite(my_ninf));
return 0;
}
#endif