mirror of
				https://github.com/0xFEEDC0DE64/arduino-esp32.git
				synced 2025-11-04 08:01:38 +01:00 
			
		
		
		
	
		
			
	
	
		
			185 lines
		
	
	
		
			7.9 KiB
		
	
	
	
		
			C
		
	
	
	
	
	
		
		
			
		
	
	
			185 lines
		
	
	
		
			7.9 KiB
		
	
	
	
		
			C
		
	
	
	
	
	
| 
								 | 
							
								/* http://pubs.opengroup.org/onlinepubs/9699919799/basedefs/tgmath.h.html */
							 | 
						||
| 
								 | 
							
								/*-
							 | 
						||
| 
								 | 
							
								 * Copyright (c) 2004 Stefan Farfeleder.
							 | 
						||
| 
								 | 
							
								 * All rights reserved.
							 | 
						||
| 
								 | 
							
								 *
							 | 
						||
| 
								 | 
							
								 * Redistribution and use in source and binary forms, with or without
							 | 
						||
| 
								 | 
							
								 * modification, are permitted provided that the following conditions
							 | 
						||
| 
								 | 
							
								 * are met:
							 | 
						||
| 
								 | 
							
								 * 1. Redistributions of source code must retain the above copyright
							 | 
						||
| 
								 | 
							
								 *    notice, this list of conditions and the following disclaimer.
							 | 
						||
| 
								 | 
							
								 * 2. Redistributions in binary form must reproduce the above copyright
							 | 
						||
| 
								 | 
							
								 *    notice, this list of conditions and the following disclaimer in the
							 | 
						||
| 
								 | 
							
								 *    documentation and/or other materials provided with the distribution.
							 | 
						||
| 
								 | 
							
								 *
							 | 
						||
| 
								 | 
							
								 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
							 | 
						||
| 
								 | 
							
								 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
							 | 
						||
| 
								 | 
							
								 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
							 | 
						||
| 
								 | 
							
								 * ARE DISCLAIMED.  IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
							 | 
						||
| 
								 | 
							
								 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
							 | 
						||
| 
								 | 
							
								 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
							 | 
						||
| 
								 | 
							
								 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
							 | 
						||
| 
								 | 
							
								 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
							 | 
						||
| 
								 | 
							
								 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
							 | 
						||
| 
								 | 
							
								 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
							 | 
						||
| 
								 | 
							
								 * SUCH DAMAGE.
							 | 
						||
| 
								 | 
							
								 *
							 | 
						||
| 
								 | 
							
								 * $FreeBSD$
							 | 
						||
| 
								 | 
							
								 */
							 | 
						||
| 
								 | 
							
								
							 | 
						||
| 
								 | 
							
								#ifndef _TGMATH_H_
							 | 
						||
| 
								 | 
							
								#define	_TGMATH_H_
							 | 
						||
| 
								 | 
							
								
							 | 
						||
| 
								 | 
							
								#include <complex.h>
							 | 
						||
| 
								 | 
							
								#include <math.h>
							 | 
						||
| 
								 | 
							
								
							 | 
						||
| 
								 | 
							
								#ifdef log2
							 | 
						||
| 
								 | 
							
								#undef log2
							 | 
						||
| 
								 | 
							
								#endif
							 | 
						||
| 
								 | 
							
								
							 | 
						||
| 
								 | 
							
								/*
							 | 
						||
| 
								 | 
							
								 * This implementation of <tgmath.h> requires two implementation-dependent
							 | 
						||
| 
								 | 
							
								 * macros to be defined:
							 | 
						||
| 
								 | 
							
								 * __tg_impl_simple(x, y, z, fn, fnf, fnl, ...)
							 | 
						||
| 
								 | 
							
								 *	Invokes fnl() if the corresponding real type of x, y or z is long
							 | 
						||
| 
								 | 
							
								 *	double, fn() if it is double or any has an integer type, and fnf()
							 | 
						||
| 
								 | 
							
								 *	otherwise.
							 | 
						||
| 
								 | 
							
								 * __tg_impl_full(x, y, z, fn, fnf, fnl, cfn, cfnf, cfnl, ...)
							 | 
						||
| 
								 | 
							
								 *	Invokes [c]fnl() if the corresponding real type of x, y or z is long
							 | 
						||
| 
								 | 
							
								 *	double, [c]fn() if it is double or any has an integer type, and
							 | 
						||
| 
								 | 
							
								 *	[c]fnf() otherwise.  The function with the 'c' prefix is called if
							 | 
						||
| 
								 | 
							
								 *	any of x, y or z is a complex number.
							 | 
						||
| 
								 | 
							
								 * Both macros call the chosen function with all additional arguments passed
							 | 
						||
| 
								 | 
							
								 * to them, as given by __VA_ARGS__.
							 | 
						||
| 
								 | 
							
								 *
							 | 
						||
| 
								 | 
							
								 * Note that these macros cannot be implemented with C's ?: operator,
							 | 
						||
| 
								 | 
							
								 * because the return type of the whole expression would incorrectly be long
							 | 
						||
| 
								 | 
							
								 * double complex regardless of the argument types.
							 | 
						||
| 
								 | 
							
								 */
							 | 
						||
| 
								 | 
							
								
							 | 
						||
| 
								 | 
							
								/* requires GCC >= 3.1 */
							 | 
						||
| 
								 | 
							
								#if !__GNUC_PREREQ (3, 1)
							 | 
						||
| 
								 | 
							
								#error "<tgmath.h> not implemented for this compiler"
							 | 
						||
| 
								 | 
							
								#endif
							 | 
						||
| 
								 | 
							
								
							 | 
						||
| 
								 | 
							
								#define	__tg_type(__e, __t)						\
							 | 
						||
| 
								 | 
							
									__builtin_types_compatible_p(__typeof__(__e), __t)
							 | 
						||
| 
								 | 
							
								#define	__tg_type3(__e1, __e2, __e3, __t)				\
							 | 
						||
| 
								 | 
							
									(__tg_type(__e1, __t) || __tg_type(__e2, __t) || 		\
							 | 
						||
| 
								 | 
							
									 __tg_type(__e3, __t))
							 | 
						||
| 
								 | 
							
								#define	__tg_type_corr(__e1, __e2, __e3, __t)				\
							 | 
						||
| 
								 | 
							
									(__tg_type3(__e1, __e2, __e3, __t) || 				\
							 | 
						||
| 
								 | 
							
									 __tg_type3(__e1, __e2, __e3, __t _Complex))
							 | 
						||
| 
								 | 
							
								#define	__tg_integer(__e1, __e2, __e3)					\
							 | 
						||
| 
								 | 
							
									(((__typeof__(__e1))1.5 == 1) || ((__typeof__(__e2))1.5 == 1) ||	\
							 | 
						||
| 
								 | 
							
									 ((__typeof__(__e3))1.5 == 1))
							 | 
						||
| 
								 | 
							
								#define	__tg_is_complex(__e1, __e2, __e3)				\
							 | 
						||
| 
								 | 
							
									(__tg_type3(__e1, __e2, __e3, float _Complex) ||		\
							 | 
						||
| 
								 | 
							
									 __tg_type3(__e1, __e2, __e3, double _Complex) ||		\
							 | 
						||
| 
								 | 
							
									 __tg_type3(__e1, __e2, __e3, long double _Complex) ||		\
							 | 
						||
| 
								 | 
							
									 __tg_type3(__e1, __e2, __e3, __typeof__(_Complex_I)))
							 | 
						||
| 
								 | 
							
								
							 | 
						||
| 
								 | 
							
								#ifdef _LDBL_EQ_DBL
							 | 
						||
| 
								 | 
							
								#define	__tg_impl_simple(x, y, z, fn, fnf, fnl, ...)			\
							 | 
						||
| 
								 | 
							
									__builtin_choose_expr(__tg_type_corr(x, y, z, long double),	\
							 | 
						||
| 
								 | 
							
									    fnl(__VA_ARGS__), __builtin_choose_expr(			\
							 | 
						||
| 
								 | 
							
										__tg_type_corr(x, y, z, double) || __tg_integer(x, y, z),\
							 | 
						||
| 
								 | 
							
										fn(__VA_ARGS__), fnf(__VA_ARGS__)))
							 | 
						||
| 
								 | 
							
								#else
							 | 
						||
| 
								 | 
							
								#define	__tg_impl_simple(__x, __y, __z, __fn, __fnf, __fnl, ...)	\
							 | 
						||
| 
								 | 
							
									(__tg_type_corr(__x, __y, __z, double) || __tg_integer(__x, __y, __z)) \
							 | 
						||
| 
								 | 
							
										? __fn(__VA_ARGS__) : __fnf(__VA_ARGS__)
							 | 
						||
| 
								 | 
							
								#endif
							 | 
						||
| 
								 | 
							
								
							 | 
						||
| 
								 | 
							
								#define	__tg_impl_full(__x, __y, __z, __fn, __fnf, __fnl, __cfn, __cfnf, __cfnl, ...)	\
							 | 
						||
| 
								 | 
							
									__builtin_choose_expr(__tg_is_complex(__x, __y, __z),		\
							 | 
						||
| 
								 | 
							
									    __tg_impl_simple(__x, __y, __z, __cfn, __cfnf, __cfnl, __VA_ARGS__),	\
							 | 
						||
| 
								 | 
							
									    __tg_impl_simple(__x, __y, __z, __fn, __fnf, __fnl, __VA_ARGS__))
							 | 
						||
| 
								 | 
							
								
							 | 
						||
| 
								 | 
							
								/* Macros to save lots of repetition below */
							 | 
						||
| 
								 | 
							
								#define	__tg_simple(__x, __fn)						\
							 | 
						||
| 
								 | 
							
									__tg_impl_simple(__x, __x, __x, __fn, __fn##f, __fn##l, __x)
							 | 
						||
| 
								 | 
							
								#define	__tg_simple2(__x, __y, __fn)					\
							 | 
						||
| 
								 | 
							
									__tg_impl_simple(__x, __x, __y, __fn, __fn##f, __fn##l, __x, __y)
							 | 
						||
| 
								 | 
							
								#define	__tg_simplev(__x, __fn, ...)					\
							 | 
						||
| 
								 | 
							
									__tg_impl_simple(__x, __x, __x, __fn, __fn##f, __fn##l, __VA_ARGS__)
							 | 
						||
| 
								 | 
							
								#define	__tg_full(__x, __fn)						\
							 | 
						||
| 
								 | 
							
									__tg_impl_full(__x, __x, __x, __fn, __fn##f, __fn##l, c##__fn, c##__fn##f, c##__fn##l, __x)
							 | 
						||
| 
								 | 
							
								
							 | 
						||
| 
								 | 
							
								/* 7.22#4 -- These macros expand to real or complex functions, depending on
							 | 
						||
| 
								 | 
							
								 * the type of their arguments. */
							 | 
						||
| 
								 | 
							
								#define	acos(__x)		__tg_full(__x, acos)
							 | 
						||
| 
								 | 
							
								#define	asin(__x)		__tg_full(__x, asin)
							 | 
						||
| 
								 | 
							
								#define	atan(__x)		__tg_full(__x, atan)
							 | 
						||
| 
								 | 
							
								#define	acosh(__x)		__tg_full(__x, acosh)
							 | 
						||
| 
								 | 
							
								#define	asinh(__x)		__tg_full(__x, asinh)
							 | 
						||
| 
								 | 
							
								#define	atanh(__x)		__tg_full(__x, atanh)
							 | 
						||
| 
								 | 
							
								#define	cos(__x)		__tg_full(__x, cos)
							 | 
						||
| 
								 | 
							
								#define	sin(__x)		__tg_full(__x, sin)
							 | 
						||
| 
								 | 
							
								#define	tan(__x)		__tg_full(__x, tan)
							 | 
						||
| 
								 | 
							
								#define	cosh(__x)		__tg_full(__x, cosh)
							 | 
						||
| 
								 | 
							
								#define	sinh(__x)		__tg_full(__x, sinh)
							 | 
						||
| 
								 | 
							
								#define	tanh(__x)		__tg_full(__x, tanh)
							 | 
						||
| 
								 | 
							
								#define	exp(__x)		__tg_full(__x, exp)
							 | 
						||
| 
								 | 
							
								#define	log(__x)		__tg_full(__x, log)
							 | 
						||
| 
								 | 
							
								#define	pow(__x, __y)		__tg_impl_full(__x, __x, __y, pow, powf, powl,	\
							 | 
						||
| 
								 | 
							
														cpow, cpowf, cpowl, __x, __y)
							 | 
						||
| 
								 | 
							
								#define	sqrt(__x)		__tg_full(__x, sqrt)
							 | 
						||
| 
								 | 
							
								
							 | 
						||
| 
								 | 
							
								/* "The corresponding type-generic macro for fabs and cabs is fabs." */
							 | 
						||
| 
								 | 
							
								#define	fabs(__x)		__tg_impl_full(__x, __x, __x, fabs, fabsf, fabsl,	\
							 | 
						||
| 
								 | 
							
														cabs, cabsf, cabsl, __x)
							 | 
						||
| 
								 | 
							
								
							 | 
						||
| 
								 | 
							
								/* 7.22#5 -- These macros are only defined for arguments with real type. */
							 | 
						||
| 
								 | 
							
								#define	atan2(__x, __y)		__tg_simple2(__x, __y, atan2)
							 | 
						||
| 
								 | 
							
								#define	cbrt(__x)		__tg_simple(__x, cbrt)
							 | 
						||
| 
								 | 
							
								#define	ceil(__x)		__tg_simple(__x, ceil)
							 | 
						||
| 
								 | 
							
								#define	copysign(__x, __y)	__tg_simple2(__x, __y, copysign)
							 | 
						||
| 
								 | 
							
								#define	erf(__x)		__tg_simple(__x, erf)
							 | 
						||
| 
								 | 
							
								#define	erfc(__x)		__tg_simple(__x, erfc)
							 | 
						||
| 
								 | 
							
								#define	exp2(__x)		__tg_simple(__x, exp2)
							 | 
						||
| 
								 | 
							
								#define	expm1(__x)		__tg_simple(__x, expm1)
							 | 
						||
| 
								 | 
							
								#define	fdim(__x, __y)		__tg_simple2(__x, __y, fdim)
							 | 
						||
| 
								 | 
							
								#define	floor(__x)		__tg_simple(__x, floor)
							 | 
						||
| 
								 | 
							
								#define	fma(__x, __y, __z)	__tg_impl_simple(__x, __y, __z, fma, fmaf, fmal, \
							 | 
						||
| 
								 | 
							
														 __x, __y, __z)
							 | 
						||
| 
								 | 
							
								#define	fmax(__x, __y)		__tg_simple2(__x, __y, fmax)
							 | 
						||
| 
								 | 
							
								#define	fmin(__x, __y)		__tg_simple2(__x, __y, fmin)
							 | 
						||
| 
								 | 
							
								#define	fmod(__x, __y)		__tg_simple2(__x, __y, fmod)
							 | 
						||
| 
								 | 
							
								#define	frexp(__x, __y)		__tg_simplev(__x, frexp, __x, __y)
							 | 
						||
| 
								 | 
							
								#define	hypot(__x, __y)		__tg_simple2(__x, __y, hypot)
							 | 
						||
| 
								 | 
							
								#define	ilogb(__x)		__tg_simple(__x, ilogb)
							 | 
						||
| 
								 | 
							
								#define	ldexp(__x, __y)		__tg_simplev(__x, ldexp, __x, __y)
							 | 
						||
| 
								 | 
							
								#define	lgamma(__x)		__tg_simple(__x, lgamma)
							 | 
						||
| 
								 | 
							
								#define	llrint(__x)		__tg_simple(__x, llrint)
							 | 
						||
| 
								 | 
							
								#define	llround(__x)		__tg_simple(__x, llround)
							 | 
						||
| 
								 | 
							
								#define	log10(__x)		__tg_simple(__x, log10)
							 | 
						||
| 
								 | 
							
								#define	log1p(__x)		__tg_simple(__x, log1p)
							 | 
						||
| 
								 | 
							
								#define	log2(__x)		__tg_simple(__x, log2)
							 | 
						||
| 
								 | 
							
								#define	logb(__x)		__tg_simple(__x, logb)
							 | 
						||
| 
								 | 
							
								#define	lrint(__x)		__tg_simple(__x, lrint)
							 | 
						||
| 
								 | 
							
								#define	lround(__x)		__tg_simple(__x, lround)
							 | 
						||
| 
								 | 
							
								#define	nearbyint(__x)		__tg_simple(__x, nearbyint)
							 | 
						||
| 
								 | 
							
								#define	nextafter(__x, __y)	__tg_simple2(__x, __y, nextafter)
							 | 
						||
| 
								 | 
							
								/* not yet implemented even for _LDBL_EQ_DBL platforms
							 | 
						||
| 
								 | 
							
								#define	nexttoward(__x, __y)	__tg_simplev(__x, nexttoward, __x, __y)
							 | 
						||
| 
								 | 
							
								*/
							 | 
						||
| 
								 | 
							
								#define	remainder(__x, __y)	__tg_simple2(__x, __y, remainder)
							 | 
						||
| 
								 | 
							
								#define	remquo(__x, __y, __z)	__tg_impl_simple(__x, __x, __y, remquo, remquof,	\
							 | 
						||
| 
								 | 
							
														 remquol, __x, __y, __z)
							 | 
						||
| 
								 | 
							
								#define	rint(__x)		__tg_simple(__x, rint)
							 | 
						||
| 
								 | 
							
								#define	round(__x)		__tg_simple(__x, round)
							 | 
						||
| 
								 | 
							
								#define	scalbn(__x, __y)	__tg_simplev(__x, scalbn, __x, __y)
							 | 
						||
| 
								 | 
							
								#define	scalbln(__x, __y)	__tg_simplev(__x, scalbln, __x, __y)
							 | 
						||
| 
								 | 
							
								#define	tgamma(__x)		__tg_simple(__x, tgamma)
							 | 
						||
| 
								 | 
							
								#define	trunc(__x)		__tg_simple(__x, trunc)
							 | 
						||
| 
								 | 
							
								
							 | 
						||
| 
								 | 
							
								/* 7.22#6 -- These macros always expand to complex functions. */
							 | 
						||
| 
								 | 
							
								#define	carg(__x)		__tg_simple(__x, carg)
							 | 
						||
| 
								 | 
							
								#define	cimag(__x)		__tg_simple(__x, cimag)
							 | 
						||
| 
								 | 
							
								#define	conj(__x)		__tg_simple(__x, conj)
							 | 
						||
| 
								 | 
							
								#define	cproj(__x)		__tg_simple(__x, cproj)
							 | 
						||
| 
								 | 
							
								#define	creal(__x)		__tg_simple(__x, creal)
							 | 
						||
| 
								 | 
							
								
							 | 
						||
| 
								 | 
							
								#endif /* !_TGMATH_H_ */
							 |