mirror of
https://git.hardenedbsd.org/hardenedbsd/HardenedBSD.git
synced 2024-11-16 07:11:05 +01:00
178 lines
5.8 KiB
C
178 lines
5.8 KiB
C
#if __GNUC__
|
|
#warning "this file includes <values.h> which is obsoleted, use <limits.h> or <float.h> instead"
|
|
#endif
|
|
|
|
/*
|
|
Copyright (C) 1988 Free Software Foundation
|
|
written by Doug Lea (dl@rocky.oswego.edu)
|
|
|
|
This file is part of the GNU C++ Library. This library is free
|
|
software; you can redistribute it and/or modify it under the terms of
|
|
the GNU Library General Public License as published by the Free
|
|
Software Foundation; either version 2 of the License, or (at your
|
|
option) any later version. This library is distributed in the hope
|
|
that it will be useful, but WITHOUT ANY WARRANTY; without even the
|
|
implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR
|
|
PURPOSE. See the GNU Library General Public License for more details.
|
|
You should have received a copy of the GNU Library General Public
|
|
License along with this library; if not, write to the Free Software
|
|
Foundation, 675 Mass Ave, Cambridge, MA 02139, USA.
|
|
*/
|
|
|
|
|
|
#ifndef _VALUES_H_
|
|
#define _VALUES_H_
|
|
|
|
#define BITSPERBYTE 8
|
|
#define BITS(type) (BITSPERBYTE * (int)sizeof(type))
|
|
|
|
#define CHARBITS BITS(char)
|
|
#define SHORTBITS BITS(short)
|
|
#define INTBITS BITS(int)
|
|
#define LONGBITS BITS(long)
|
|
#define PTRBITS BITS(char*)
|
|
#define DOUBLEBITS BITS(double)
|
|
#define FLOATBITS BITS(float)
|
|
|
|
#define MINSHORT ((short)(1 << (SHORTBITS - 1)))
|
|
#define MININT (1 << (INTBITS - 1))
|
|
#define MINLONG (1L << (LONGBITS - 1))
|
|
|
|
#define MAXSHORT ((short)~MINSHORT)
|
|
#define MAXINT (~MININT)
|
|
#define MAXLONG (~MINLONG)
|
|
|
|
#define HIBITS MINSHORT
|
|
#define HIBITL MINLONG
|
|
|
|
#if defined(sun) || defined(hp300) || defined(hpux) || defined(masscomp) || defined(sgi)
|
|
#ifdef masscomp
|
|
#define MAXDOUBLE \
|
|
({ \
|
|
double maxdouble_val; \
|
|
\
|
|
__asm ("fmove%.d #0x7fefffffffffffff,%0" /* Max double */ \
|
|
: "=f" (maxdouble_val) \
|
|
: /* no inputs */); \
|
|
maxdouble_val; \
|
|
})
|
|
#define MAXFLOAT ((float) 3.40e+38)
|
|
#else
|
|
#define MAXDOUBLE 1.79769313486231470e+308
|
|
#define MAXFLOAT ((float)3.40282346638528860e+38)
|
|
#endif
|
|
#define MINDOUBLE 4.94065645841246544e-324
|
|
#define MINFLOAT ((float)1.40129846432481707e-45)
|
|
#define _IEEE 1
|
|
#define _DEXPLEN 11
|
|
#define _FEXPLEN 8
|
|
#define _HIDDENBIT 1
|
|
#define DMINEXP (-(DMAXEXP + DSIGNIF - _HIDDENBIT - 3))
|
|
#define FMINEXP (-(FMAXEXP + FSIGNIF - _HIDDENBIT - 3))
|
|
#define DMAXEXP ((1 << _DEXPLEN - 1) - 1 + _IEEE)
|
|
#define FMAXEXP ((1 << _FEXPLEN - 1) - 1 + _IEEE)
|
|
|
|
#elif defined(sony)
|
|
#define MAXDOUBLE 1.79769313486231470e+308
|
|
#define MAXFLOAT ((float)3.40282346638528860e+38)
|
|
#define MINDOUBLE 2.2250738585072010e-308
|
|
#define MINFLOAT ((float)1.17549435e-38)
|
|
#define _IEEE 1
|
|
#define _DEXPLEN 11
|
|
#define _FEXPLEN 8
|
|
#define _HIDDENBIT 1
|
|
#define DMINEXP (-(DMAXEXP + DSIGNIF - _HIDDENBIT - 3))
|
|
#define FMINEXP (-(FMAXEXP + FSIGNIF - _HIDDENBIT - 3))
|
|
#define DMAXEXP ((1 << _DEXPLEN - 1) - 1 + _IEEE)
|
|
#define FMAXEXP ((1 << _FEXPLEN - 1) - 1 + _IEEE)
|
|
|
|
#elif defined(sequent)
|
|
extern double _maxdouble, _mindouble;
|
|
extern float _maxfloat, _minfloat;
|
|
#define MAXDOUBLE _maxdouble
|
|
#define MAXFLOAT _maxfloat
|
|
#define MINDOUBLE _mindouble
|
|
#define MINFLOAT _minfloat
|
|
#define _IEEE 1
|
|
#define _DEXPLEN 11
|
|
#define _FEXPLEN 8
|
|
#define _HIDDENBIT 1
|
|
#define DMINEXP (-(DMAXEXP - 3))
|
|
#define FMINEXP (-(FMAXEXP - 3))
|
|
#define DMAXEXP ((1 << _DEXPLEN - 1) - 1 + _IEEE)
|
|
#define FMAXEXP ((1 << _FEXPLEN - 1) - 1 + _IEEE)
|
|
|
|
#elif defined(i386)
|
|
#define MAXDOUBLE 1.79769313486231570e+308
|
|
#define MAXFLOAT ((float)3.40282346638528860e+38)
|
|
#define MINDOUBLE 2.22507385850720140e-308
|
|
#define MINFLOAT ((float)1.17549435082228750e-38)
|
|
#define _IEEE 0
|
|
#define _DEXPLEN 11
|
|
#define _FEXPLEN 8
|
|
#define _HIDDENBIT 1
|
|
#define DMINEXP (-DMAXEXP)
|
|
#define FMINEXP (-FMAXEXP)
|
|
#define DMAXEXP ((1 << _DEXPLEN - 1) - 1 + _IEEE)
|
|
#define FMAXEXP ((1 << _FEXPLEN - 1) - 1 + _IEEE)
|
|
|
|
/* from Andrew Klossner <andrew%frip.wv.tek.com@relay.cs.net> */
|
|
#elif defined(m88k)
|
|
/* These are "good" guesses ...
|
|
I'll figure out the true mins and maxes later, at the time I find
|
|
out the mins and maxes that the compiler can tokenize. */
|
|
#define MAXDOUBLE 1.79769313486231e+308
|
|
#define MAXFLOAT ((float)3.40282346638528e+38)
|
|
#define MINDOUBLE 2.22507385850720e-308
|
|
#define MINFLOAT ((float)1.17549435082228e-38)
|
|
#define _IEEE 1
|
|
#define _DEXPLEN 11
|
|
#define _FEXPLEN 8
|
|
#define _HIDDENBIT 1
|
|
#define DMINEXP (1-DMAXEXP)
|
|
#define FMINEXP (1-FMAXEXP)
|
|
#define DMAXEXP ((1 << _DEXPLEN - 1) - 1 + _IEEE)
|
|
#define FMAXEXP ((1 << _FEXPLEN - 1) - 1 + _IEEE)
|
|
|
|
#elif defined(convex)
|
|
#define MAXDOUBLE 8.9884656743115785e+306
|
|
#define MAXFLOAT ((float) 1.70141173e+38)
|
|
#define MINDOUBLE 5.5626846462680035e-308
|
|
#define MINFLOAT ((float) 2.93873588e-39)
|
|
#define _IEEE 0
|
|
#define _DEXPLEN 11
|
|
#define _FEXPLEN 8
|
|
#define _HIDDENBIT 1
|
|
#define DMINEXP (-DMAXEXP)
|
|
#define FMINEXP (-FMAXEXP)
|
|
#define DMAXEXP ((1 << _DEXPLEN - 1) - 1 + _IEEE)
|
|
#define FMAXEXP ((1 << _FEXPLEN - 1) - 1 + _IEEE)
|
|
|
|
/* #elif defined(vax) */
|
|
/* use vax versions by default -- they seem to be the most conservative */
|
|
#else
|
|
|
|
#define MAXDOUBLE 1.701411834604692293e+38
|
|
#define MINDOUBLE (2.938735877055718770e-39)
|
|
|
|
#define MAXFLOAT 1.7014117331926443e+38
|
|
#define MINFLOAT 2.9387358770557188e-39
|
|
|
|
#define _IEEE 0
|
|
#define _DEXPLEN 8
|
|
#define _FEXPLEN 8
|
|
#define _HIDDENBIT 1
|
|
#define DMINEXP (-DMAXEXP)
|
|
#define FMINEXP (-FMAXEXP)
|
|
#define DMAXEXP ((1 << _DEXPLEN - 1) - 1 + _IEEE)
|
|
#define FMAXEXP ((1 << _FEXPLEN - 1) - 1 + _IEEE)
|
|
#endif
|
|
|
|
#define DSIGNIF (DOUBLEBITS - _DEXPLEN + _HIDDENBIT - 1)
|
|
#define FSIGNIF (FLOATBITS - _FEXPLEN + _HIDDENBIT - 1)
|
|
#define DMAXPOWTWO ((double)(1L << LONGBITS -2)*(1L << DSIGNIF - LONGBITS +1))
|
|
#define FMAXPOWTWO ((float)(1L << FSIGNIF - 1))
|
|
|
|
#endif /* !_VALUES_H_ */
|
|
|