aboutsummaryrefslogtreecommitdiff
path: root/math
diff options
context:
space:
mode:
authorUlrich Drepper <drepper@redhat.com>2002-08-28 08:01:11 +0000
committerUlrich Drepper <drepper@redhat.com>2002-08-28 08:01:11 +0000
commit7a5affebe810c786bfc9ef17cdc4d0bc20a1302e (patch)
treed461b7135c42ab5555bddfe4a5fa4b46a1c4ca94 /math
parenta3f9038c879c11344430169d6daa8fad30d4f379 (diff)
downloadglibc-7a5affebe810c786bfc9ef17cdc4d0bc20a1302e.tar
glibc-7a5affebe810c786bfc9ef17cdc4d0bc20a1302e.tar.gz
glibc-7a5affebe810c786bfc9ef17cdc4d0bc20a1302e.tar.bz2
glibc-7a5affebe810c786bfc9ef17cdc4d0bc20a1302e.zip
Prepare headers for use in ISO C++ compliant implementations.
Diffstat (limited to 'math')
-rw-r--r--math/bits/mathcalls.h69
-rw-r--r--math/math.h14
2 files changed, 62 insertions, 21 deletions
diff --git a/math/bits/mathcalls.h b/math/bits/mathcalls.h
index 12fcfce022..59d4eb88f7 100644
--- a/math/bits/mathcalls.h
+++ b/math/bits/mathcalls.h
@@ -1,5 +1,5 @@
/* Prototype declarations for math functions; helper file for <math.h>.
- Copyright (C) 1996,1997,1998,1999,2000,2001 Free Software Foundation, Inc.
+ Copyright (C) 1996-2001, 2002 Free Software Foundation, Inc.
This file is part of the GNU C Library.
The GNU C Library is free software; you can redistribute it and/or
@@ -44,12 +44,13 @@
NAME, to make token pasting work with -traditional. */
#ifndef _MATH_H
- #error "Never include <bits/mathcalls.h> directly; include <math.h> instead."
+# error "Never include <bits/mathcalls.h> directly; include <math.h> instead."
#endif
/* Trigonometric functions. */
+_Mdouble_BEGIN_NAMESPACE
/* Arc cosine of X. */
__MATHCALL (acos,, (_Mdouble_ __x));
/* Arc sine of X. */
@@ -66,12 +67,6 @@ __MATHCALL (sin,, (_Mdouble_ __x));
/* Tangent of X. */
__MATHCALL (tan,, (_Mdouble_ __x));
-#ifdef __USE_GNU
-/* Cosine and sine of X. */
-__MATHDECL (void,sincos,,
- (_Mdouble_ __x, _Mdouble_ *__sinx, _Mdouble_ *__cosx));
-#endif
-
/* Hyperbolic functions. */
/* Hyperbolic cosine of X. */
@@ -80,28 +75,31 @@ __MATHCALL (cosh,, (_Mdouble_ __x));
__MATHCALL (sinh,, (_Mdouble_ __x));
/* Hyperbolic tangent of X. */
__MATHCALL (tanh,, (_Mdouble_ __x));
+_Mdouble_END_NAMESPACE
+
+#ifdef __USE_GNU
+/* Cosine and sine of X. */
+__MATHDECL (void,sincos,,
+ (_Mdouble_ __x, _Mdouble_ *__sinx, _Mdouble_ *__cosx));
+#endif
#if defined __USE_MISC || defined __USE_XOPEN_EXTENDED || defined __USE_ISOC99
+__BEGIN_NAMESPACE_C99
/* Hyperbolic arc cosine of X. */
__MATHCALL (acosh,, (_Mdouble_ __x));
/* Hyperbolic arc sine of X. */
__MATHCALL (asinh,, (_Mdouble_ __x));
/* Hyperbolic arc tangent of X. */
__MATHCALL (atanh,, (_Mdouble_ __x));
+__END_NAMESPACE_C99
#endif
/* Exponential and logarithmic functions. */
+_Mdouble_BEGIN_NAMESPACE
/* Exponential function of X. */
__MATHCALL (exp,, (_Mdouble_ __x));
-#ifdef __USE_GNU
-/* A function missing in all standards: compute exponent to base ten. */
-__MATHCALL (exp10,, (_Mdouble_ __x));
-/* Another name occasionally used. */
-__MATHCALL (pow10,, (_Mdouble_ __x));
-#endif
-
/* Break VALUE into a normalized fraction and an integral power of 2. */
__MATHCALL (frexp,, (_Mdouble_ __x, int *__exponent));
@@ -116,8 +114,17 @@ __MATHCALL (log10,, (_Mdouble_ __x));
/* Break VALUE into integral and fractional parts. */
__MATHCALL (modf,, (_Mdouble_ __x, _Mdouble_ *__iptr));
+_Mdouble_END_NAMESPACE
+
+#ifdef __USE_GNU
+/* A function missing in all standards: compute exponent to base ten. */
+__MATHCALL (exp10,, (_Mdouble_ __x));
+/* Another name occasionally used. */
+__MATHCALL (pow10,, (_Mdouble_ __x));
+#endif
#if defined __USE_MISC || defined __USE_XOPEN_EXTENDED || defined __USE_ISOC99
+__BEGIN_NAMESPACE_C99
/* Return exp(X) - 1. */
__MATHCALL (expm1,, (_Mdouble_ __x));
@@ -126,38 +133,48 @@ __MATHCALL (log1p,, (_Mdouble_ __x));
/* Return the base 2 signed integral exponent of X. */
__MATHCALL (logb,, (_Mdouble_ __x));
+__END_NAMESPACE_C99
#endif
#ifdef __USE_ISOC99
+__BEGIN_NAMESPACE_C99
/* Compute base-2 exponential of X. */
__MATHCALL (exp2,, (_Mdouble_ __x));
/* Compute base-2 logarithm of X. */
__MATHCALL (log2,, (_Mdouble_ __x));
+__END_NAMESPACE_C99
#endif
/* Power functions. */
+_Mdouble_BEGIN_NAMESPACE
/* Return X to the Y power. */
__MATHCALL (pow,, (_Mdouble_ __x, _Mdouble_ __y));
/* Return the square root of X. */
__MATHCALL (sqrt,, (_Mdouble_ __x));
+_Mdouble_END_NAMESPACE
#if defined __USE_MISC || defined __USE_XOPEN || defined __USE_ISOC99
+__BEGIN_NAMESPACE_C99
/* Return `sqrt(X*X + Y*Y)'. */
__MATHCALL (hypot,, (_Mdouble_ __x, _Mdouble_ __y));
+__END_NAMESPACE_C99
#endif
#if defined __USE_MISC || defined __USE_XOPEN_EXTENDED || defined __USE_ISOC99
+__BEGIN_NAMESPACE_C99
/* Return the cube root of X. */
__MATHCALL (cbrt,, (_Mdouble_ __x));
+__END_NAMESPACE_C99
#endif
/* Nearest integer, absolute value, and remainder functions. */
+_Mdouble_BEGIN_NAMESPACE
/* Smallest integral value not less than X. */
__MATHCALLX (ceil,, (_Mdouble_ __x), (__const__));
@@ -177,6 +194,7 @@ __MATHDECL_1 (int,__isinf,, (_Mdouble_ __value)) __attribute__ ((__const__));
/* Return nonzero if VALUE is finite and not NaN. */
__MATHDECL_1 (int,__finite,, (_Mdouble_ __value)) __attribute__ ((__const__));
+_Mdouble_END_NAMESPACE
#ifdef __USE_MISC
/* Return 0 if VALUE is finite or NaN, +1 if it
@@ -195,13 +213,17 @@ __MATHCALL (significand,, (_Mdouble_ __x));
#endif /* Use misc. */
#if defined __USE_MISC || defined __USE_ISOC99
+__BEGIN_NAMESPACE_C99
/* Return X with its signed changed to Y's. */
__MATHCALLX (copysign,, (_Mdouble_ __x, _Mdouble_ __y), (__const__));
+__END_NAMESPACE_C99
#endif
#ifdef __USE_ISOC99
+__BEGIN_NAMESPACE_C99
/* Return representation of NaN for double type. */
__MATHCALLX (nan,, (__const char *__tagb), (__const__));
+__END_NAMESPACE_C99
#endif
@@ -223,14 +245,19 @@ __MATHCALL (yn,, (int, _Mdouble_));
#if defined __USE_MISC || defined __USE_XOPEN || defined __USE_ISOC99
+__BEGIN_NAMESPACE_C99
/* Error and gamma functions. */
__MATHCALL (erf,, (_Mdouble_));
__MATHCALL (erfc,, (_Mdouble_));
__MATHCALL (lgamma,, (_Mdouble_));
+__END_NAMESPACE_C99
#endif
#ifdef __USE_ISOC99
+__BEGIN_NAMESPACE_C99
+/* True gamma function. */
__MATHCALL (tgamma,, (_Mdouble_));
+__END_NAMESPACE_C99
#endif
#if defined __USE_MISC || defined __USE_XOPEN
@@ -247,6 +274,7 @@ __MATHCALL (lgamma,_r, (_Mdouble_, int *__signgamp));
#if defined __USE_MISC || defined __USE_XOPEN_EXTENDED || defined __USE_ISOC99
+__BEGIN_NAMESPACE_C99
/* Return the integer nearest X in the direction of the
prevailing rounding mode. */
__MATHCALL (rint,, (_Mdouble_ __x));
@@ -260,11 +288,6 @@ __MATHCALLX (nexttoward,, (_Mdouble_ __x, long double __y), (__const__));
/* Return the remainder of integer divison X / Y with infinite precision. */
__MATHCALL (remainder,, (_Mdouble_ __x, _Mdouble_ __y));
-# if defined __USE_MISC || defined __USE_XOPEN_EXTENDED
-/* Return X times (2 to the Nth power). */
-__MATHCALL (scalb,, (_Mdouble_ __x, _Mdouble_ __n));
-# endif
-
# if defined __USE_MISC || defined __USE_ISOC99
/* Return X times (2 to the Nth power). */
__MATHCALL (scalbn,, (_Mdouble_ __x, int __n));
@@ -330,4 +353,10 @@ __MATHDECL_1 (int, __signbit,, (_Mdouble_ __value))
/* Multiply-add function computed as a ternary operation. */
__MATHCALL (fma,, (_Mdouble_ __x, _Mdouble_ __y, _Mdouble_ __z));
+__END_NAMESPACE_C99
+
+# if defined __USE_MISC || defined __USE_XOPEN_EXTENDED
+/* Return X times (2 to the Nth power). */
+__MATHCALL (scalb,, (_Mdouble_ __x, _Mdouble_ __n));
+# endif
#endif /* Use ISO C99. */
diff --git a/math/math.h b/math/math.h
index 8a2e86c2f1..28e5937c56 100644
--- a/math/math.h
+++ b/math/math.h
@@ -1,5 +1,5 @@
/* Declarations for math functions.
- Copyright (C) 1991,92,93,95,96,97,98,99,2001 Free Software Foundation, Inc.
+ Copyright (C) 1991-1993,1995-1999,2001,2002 Free Software Foundation, Inc.
This file is part of the GNU C Library.
The GNU C Library is free software; you can redistribute it and/or
@@ -60,8 +60,12 @@ __BEGIN_DECLS
#define _Mdouble_ double
#define __MATH_PRECNAME(name,r) __CONCAT(name,r)
+# define _Mdouble_BEGIN_NAMESPACE __BEGIN_NAMESPACE_STD
+# define _Mdouble_END_NAMESPACE __END_NAMESPACE_STD
#include <bits/mathcalls.h>
#undef _Mdouble_
+#undef _Mdouble_BEGIN_NAMESPACE
+#undef _Mdouble_END_NAMESPACE
#undef __MATH_PRECNAME
#if defined __USE_MISC || defined __USE_ISOC99
@@ -79,8 +83,12 @@ __BEGIN_DECLS
# else
# define __MATH_PRECNAME(name,r) name/**/f/**/r
# endif
+# define _Mdouble_BEGIN_NAMESPACE __BEGIN_NAMESPACE_C99
+# define _Mdouble_END_NAMESPACE __END_NAMESPACE_C99
# include <bits/mathcalls.h>
# undef _Mdouble_
+# undef _Mdouble_BEGIN_NAMESPACE
+# undef _Mdouble_END_NAMESPACE
# undef __MATH_PRECNAME
# if (__STDC__ - 0 || __GNUC__ - 0) && !defined __NO_LONG_DOUBLE_MATH
@@ -96,8 +104,12 @@ __BEGIN_DECLS
# else
# define __MATH_PRECNAME(name,r) name/**/l/**/r
# endif
+# define _Mdouble_BEGIN_NAMESPACE __BEGIN_NAMESPACE_C99
+# define _Mdouble_END_NAMESPACE __END_NAMESPACE_C99
# include <bits/mathcalls.h>
# undef _Mdouble_
+# undef _Mdouble_BEGIN_NAMESPACE
+# undef _Mdouble_END_NAMESPACE
# undef __MATH_PRECNAME
# endif /* __STDC__ || __GNUC__ */