gen-num-limits.cc (traps): Add fflush prior to trapping signal (UnixWare 7 workaround).
2001-03-12 Rodney Brown <RodneyBrown@mynd.com>
* src/gen-num-limits.cc (traps) [SPECIALIZE_TRAPPING]: Add fflush
prior to trapping signal (UnixWare 7 workaround).
From-SVN: r40422
This commit is contained in:
Rodney Brown
Benjamin Kosnik
parent
23712c118d
commit
e6ff97b1c4
|
|
@ -1,3 +1,8 @@
|
|||
2001-03-12 Rodney Brown <RodneyBrown@mynd.com>
|
||||
|
||||
* src/gen-num-limits.cc (traps) [SPECIALIZE_TRAPPING]: Add fflush
|
||||
prior to trapping signal (UnixWare 7 workaround).
|
||||
|
||||
2001-03-12 Craig Rodrigues <rodrigc@mediaone.net>
|
||||
|
||||
* include/backward/hashtable.h: Same.
|
||||
|
|
|
|||
|
|
@ -139,49 +139,55 @@ bool trapping(const Operation& op)
|
|||
return false;
|
||||
}
|
||||
|
||||
template<typename T> struct division_by_zero {
|
||||
template<typename T>
|
||||
struct division_by_zero
|
||||
{
|
||||
void operator() () const
|
||||
{
|
||||
volatile T zero = T();
|
||||
volatile T one = T(1);
|
||||
volatile T infinity = one / zero;
|
||||
volatile T zero = T();
|
||||
volatile T one = T(1);
|
||||
volatile T infinity = one / zero;
|
||||
}
|
||||
};
|
||||
};
|
||||
|
||||
template<typename T> struct overflow {
|
||||
template<typename T>
|
||||
struct overflow
|
||||
{
|
||||
void operator() () const
|
||||
{
|
||||
T i = T(1);
|
||||
T j = T();
|
||||
while (i>j) {
|
||||
j = i;
|
||||
i = i * 2 + 1;
|
||||
}
|
||||
T i = T(1);
|
||||
T j = T();
|
||||
while (i>j) {
|
||||
j = i;
|
||||
i = i * 2 + 1;
|
||||
}
|
||||
}
|
||||
};
|
||||
};
|
||||
|
||||
template<typename T> struct underflow {};
|
||||
template<typename T> struct underflow { };
|
||||
|
||||
// traps
|
||||
template<typename T> void traps()
|
||||
{
|
||||
signal_adapter (signal, SIGFPE, signal_handler);
|
||||
signal_adapter (signal, SIGTRAP, signal_handler);
|
||||
bool trap_flag = trapping(division_by_zero<T>());
|
||||
signal_adapter (signal, SIGFPE, signal_handler);
|
||||
signal_adapter (signal, SIGTRAP, signal_handler);
|
||||
trap_flag = trap_flag || trapping(overflow<T>());
|
||||
const char* p = bool_alpha[trap_flag];
|
||||
printf("%s%s = %s;\n", tab2, "static const bool traps", p);
|
||||
fflush(NULL);
|
||||
signal_adapter (signal, SIGFPE, signal_handler);
|
||||
signal_adapter (signal, SIGTRAP, signal_handler);
|
||||
bool trap_flag = trapping(division_by_zero<T>());
|
||||
signal_adapter (signal, SIGFPE, signal_handler);
|
||||
signal_adapter (signal, SIGTRAP, signal_handler);
|
||||
trap_flag = trap_flag || trapping(overflow<T>());
|
||||
const char* p = bool_alpha[trap_flag];
|
||||
printf("%s%s = %s;\n", tab2, "static const bool traps", p);
|
||||
}
|
||||
|
||||
#define SPECIALIZE_TRAPPING(T) \
|
||||
template<> void traps< T >() \
|
||||
{ \
|
||||
signal_adapter (signal, SIGFPE, signal_handler); \
|
||||
signal_adapter (signal, SIGTRAP, signal_handler); \
|
||||
const char* p = bool_alpha[trapping(division_by_zero<T>())]; \
|
||||
printf("%s%s = %s;\n", tab2, "static const bool traps", p); \
|
||||
{ \
|
||||
fflush(NULL); \
|
||||
signal_adapter (signal, SIGFPE, signal_handler); \
|
||||
signal_adapter (signal, SIGTRAP, signal_handler); \
|
||||
const char* p = bool_alpha[trapping(division_by_zero<T>())]; \
|
||||
printf("%s%s = %s;\n", tab2, "static const bool traps", p); \
|
||||
}
|
||||
|
||||
SPECIALIZE_TRAPPING(unsigned char);
|
||||
|
|
@ -194,10 +200,12 @@ SPECIALIZE_TRAPPING(unsigned long long);
|
|||
|
||||
#undef SPECIALIZE_TRAPPING
|
||||
|
||||
template<typename T> struct type_name_trait {
|
||||
template<typename T>
|
||||
struct type_name_trait
|
||||
{
|
||||
static const char type_name[];
|
||||
static const char trait_name[];
|
||||
};
|
||||
};
|
||||
|
||||
#define DEFINED_TYPE_NAME(T) \
|
||||
const char type_name_trait< T >::type_name[] = #T; \
|
||||
|
|
@ -225,40 +233,41 @@ DEFINED_TYPE_NAME(long double);
|
|||
#undef DEFINED_TYPE_NAME
|
||||
|
||||
// declarator
|
||||
template<typename T> struct declarator : type_name_trait<T> {
|
||||
template<typename T>
|
||||
struct declarator : type_name_trait<T>
|
||||
{
|
||||
typedef type_name_trait<T> base;
|
||||
static void start()
|
||||
{
|
||||
printf("%s%s %s %s\n", tab, "template<> struct",
|
||||
base::trait_name, "{");
|
||||
{
|
||||
printf("%s%s %s %s\n", tab, "template<> struct", base::trait_name, "{");
|
||||
}
|
||||
|
||||
|
||||
static void end()
|
||||
{
|
||||
printf("%s};\n\n", tab);
|
||||
}
|
||||
};
|
||||
{ printf("%s};\n\n", tab); }
|
||||
};
|
||||
|
||||
|
||||
//
|
||||
// Predicates
|
||||
//
|
||||
template<typename T> struct predicate {
|
||||
template<typename T>
|
||||
struct predicate
|
||||
{
|
||||
static const bool is_signed;
|
||||
static const bool is_integer;
|
||||
static const bool is_exact;
|
||||
|
||||
|
||||
static const bool has_infinity;
|
||||
static const bool has_quiet_nan;
|
||||
static const bool has_signaling_nan;
|
||||
static const bool has_denorm;
|
||||
static const bool has_denorm_loss;
|
||||
|
||||
|
||||
static const bool is_iec559;
|
||||
static const bool is_bounded;
|
||||
|
||||
|
||||
static const bool traps;
|
||||
};
|
||||
};
|
||||
|
||||
template<typename T>
|
||||
const bool predicate<T>::is_signed = T(-1) < 0;
|
||||
|
|
@ -321,23 +330,24 @@ SPECIALIZE_IEC559(unsigned long long);
|
|||
//
|
||||
// Values
|
||||
//
|
||||
|
||||
template<typename T> struct value {
|
||||
template<typename T>
|
||||
struct value
|
||||
{
|
||||
static const char min[];
|
||||
static const char max[];
|
||||
|
||||
|
||||
static const int digits;
|
||||
static const int digits10;
|
||||
|
||||
static const int radix;
|
||||
static const char epsilon[];
|
||||
static const char round_error[];
|
||||
|
||||
|
||||
static const int min_exponent;
|
||||
static const int min_exponent10;
|
||||
static const int max_exponent;
|
||||
static const int max_exponent10;
|
||||
};
|
||||
};
|
||||
|
||||
#define DEFINE_EXTREMA(T, m, M) DO_DEFINE_EXTREMA(T, m, M)
|
||||
#define DO_DEFINE_EXTREMA(T, m, M) \
|
||||
|
|
@ -507,283 +517,285 @@ SPECIALIZE_EXPONENTS(long double, LDBL_MIN_EXP, LDBL_MIN_10_EXP,
|
|||
|
||||
template<typename T> void is_signed()
|
||||
{
|
||||
printf("%s%s = %s;\n", tab2, "static const bool is_signed",
|
||||
bool_alpha[predicate<T>::is_signed]);
|
||||
printf("%s%s = %s;\n", tab2, "static const bool is_signed",
|
||||
bool_alpha[predicate<T>::is_signed]);
|
||||
}
|
||||
|
||||
// a fundamental type is modulo iff it isn't signed
|
||||
template<typename T> void is_modulo()
|
||||
{
|
||||
printf("%s%s = %s;\n", tab2, "static const bool is_modulo",
|
||||
bool_alpha[! predicate<T>::is_signed]);
|
||||
printf("%s%s = %s;\n", tab2, "static const bool is_modulo",
|
||||
bool_alpha[! predicate<T>::is_signed]);
|
||||
}
|
||||
|
||||
template<typename T>
|
||||
void min()
|
||||
{
|
||||
printf("%s%s%s%s\n%s%s%s%s\n", tab2, "static ", declarator<T>::type_name,
|
||||
" min() throw()", tab2, "{ return ", value<T>::min, "; }");
|
||||
printf("%s%s%s%s\n%s%s%s%s\n", tab2, "static ", declarator<T>::type_name,
|
||||
" min() throw()", tab2, "{ return ", value<T>::min, "; }");
|
||||
}
|
||||
|
||||
template<typename T>
|
||||
void max()
|
||||
{
|
||||
printf("%s%s%s%s\n%s%s%s%s\n", tab2, "static ", declarator<T>::type_name,
|
||||
" max() throw()", tab2, "{ return ", value<T>::max, "; }");
|
||||
printf("%s%s%s%s\n%s%s%s%s\n", tab2, "static ", declarator<T>::type_name,
|
||||
" max() throw()", tab2, "{ return ", value<T>::max, "; }");
|
||||
}
|
||||
|
||||
template<typename T>
|
||||
void is_integer()
|
||||
{
|
||||
printf("%s%s = %s;\n", tab2, "static const bool is_integer",
|
||||
bool_alpha[predicate<T>::is_integer]);
|
||||
printf("%s%s = %s;\n", tab2, "static const bool is_integer",
|
||||
bool_alpha[predicate<T>::is_integer]);
|
||||
}
|
||||
|
||||
template<typename T>
|
||||
void is_exact()
|
||||
{
|
||||
printf("%s%s = %s;\n", tab2, "static const bool is_exact",
|
||||
bool_alpha[predicate<T>::is_exact]);
|
||||
printf("%s%s = %s;\n", tab2, "static const bool is_exact",
|
||||
bool_alpha[predicate<T>::is_exact]);
|
||||
}
|
||||
|
||||
template<typename T>
|
||||
void digits()
|
||||
{
|
||||
printf("%s%s = %d;\n", tab2, "static const int digits",
|
||||
value<T>::digits);
|
||||
printf("%s%s = %d;\n", tab2, "static const int digits",
|
||||
value<T>::digits);
|
||||
}
|
||||
|
||||
template<typename T>
|
||||
void digits10()
|
||||
{
|
||||
printf("%s%s = %d;\n", tab2, "static const int digits10",
|
||||
int(log10_of_two * value<T>::digits));
|
||||
printf("%s%s = %d;\n", tab2, "static const int digits10",
|
||||
int(log10_of_two * value<T>::digits));
|
||||
}
|
||||
|
||||
template<typename T>
|
||||
void radix()
|
||||
{
|
||||
printf("%s%s = %d;\n", tab2, "static const int radix",
|
||||
value<T>::radix);
|
||||
printf("%s%s = %d;\n", tab2, "static const int radix",
|
||||
value<T>::radix);
|
||||
}
|
||||
|
||||
template<typename T>
|
||||
void epsilon()
|
||||
{
|
||||
printf("%s%s %s %s\n%s%s %s%s\n", tab2, "static",
|
||||
declarator<T>::type_name, "epsilon() throw()",
|
||||
tab2, "{ return", value<T>::epsilon, "; }");
|
||||
printf("%s%s %s %s\n%s%s %s%s\n", tab2, "static",
|
||||
declarator<T>::type_name, "epsilon() throw()",
|
||||
tab2, "{ return", value<T>::epsilon, "; }");
|
||||
}
|
||||
|
||||
template<typename T>
|
||||
void round_error()
|
||||
{
|
||||
printf("%s%s %s %s\n%s%s %s%s\n", tab2, "static",
|
||||
declarator<T>::type_name, "round_error() throw()",
|
||||
tab2, "{ return", value<T>::round_error, "; }");
|
||||
printf("%s%s %s %s\n%s%s %s%s\n", tab2, "static",
|
||||
declarator<T>::type_name, "round_error() throw()",
|
||||
tab2, "{ return", value<T>::round_error, "; }");
|
||||
}
|
||||
|
||||
template<typename T>
|
||||
void min_exponent()
|
||||
{
|
||||
printf("%s%s = %d;\n", tab2, "static const int min_exponent",
|
||||
value<T>::min_exponent);
|
||||
printf("%s%s = %d;\n", tab2, "static const int min_exponent",
|
||||
value<T>::min_exponent);
|
||||
}
|
||||
|
||||
template<typename T>
|
||||
void min_exponent10()
|
||||
{
|
||||
printf("%s%s = %d;\n", tab2, "static const int min_exponent10",
|
||||
value<T>::min_exponent10);
|
||||
printf("%s%s = %d;\n", tab2, "static const int min_exponent10",
|
||||
value<T>::min_exponent10);
|
||||
}
|
||||
|
||||
template<typename T>
|
||||
void max_exponent()
|
||||
{
|
||||
printf("%s%s = %d;\n", tab2, "static const int max_exponent",
|
||||
value<T>::max_exponent);
|
||||
printf("%s%s = %d;\n", tab2, "static const int max_exponent",
|
||||
value<T>::max_exponent);
|
||||
}
|
||||
|
||||
template<typename T>
|
||||
void max_exponent10()
|
||||
{
|
||||
printf("%s%s = %d;\n", tab2, "static const int max_exponent10",
|
||||
value<T>::max_exponent10);
|
||||
printf("%s%s = %d;\n", tab2, "static const int max_exponent10",
|
||||
value<T>::max_exponent10);
|
||||
}
|
||||
|
||||
template<typename T>
|
||||
void has_infinity()
|
||||
{
|
||||
printf("%s%s = %s;\n", tab2, "static const bool has_infinity",
|
||||
bool_alpha[predicate<T>::has_infinity]);
|
||||
printf("%s%s = %s;\n", tab2, "static const bool has_infinity",
|
||||
bool_alpha[predicate<T>::has_infinity]);
|
||||
}
|
||||
|
||||
template<typename T>
|
||||
void has_quiet_nan()
|
||||
{
|
||||
printf("%s%s = %s;\n", tab2, "static const bool has_quiet_NaN",
|
||||
bool_alpha[predicate<T>::has_quiet_nan]);
|
||||
printf("%s%s = %s;\n", tab2, "static const bool has_quiet_NaN",
|
||||
bool_alpha[predicate<T>::has_quiet_nan]);
|
||||
}
|
||||
|
||||
template<typename T>
|
||||
void has_signaling_nan()
|
||||
{
|
||||
printf("%s%s = %s;\n", tab2, "static const bool has_signaling_NaN",
|
||||
bool_alpha[predicate<T>::has_signaling_nan]);
|
||||
printf("%s%s = %s;\n", tab2, "static const bool has_signaling_NaN",
|
||||
bool_alpha[predicate<T>::has_signaling_nan]);
|
||||
}
|
||||
|
||||
template<typename T>
|
||||
void has_denorm_loss()
|
||||
{
|
||||
printf("%s%s = %s;\n", tab2, "static const bool has_denorm_loss",
|
||||
bool_alpha[predicate<T>::has_denorm_loss]);
|
||||
printf("%s%s = %s;\n", tab2, "static const bool has_denorm_loss",
|
||||
bool_alpha[predicate<T>::has_denorm_loss]);
|
||||
}
|
||||
|
||||
template<typename T> struct infinity_trait {
|
||||
template<typename T>
|
||||
struct infinity_trait
|
||||
{
|
||||
static void has_denorm()
|
||||
{
|
||||
printf("%s%s;\n", tab2, "static const float_denorm_style "
|
||||
"has_denorm = denorm_absent");
|
||||
printf("%s%s;\n", tab2, "static const float_denorm_style "
|
||||
"has_denorm = denorm_absent");
|
||||
}
|
||||
|
||||
|
||||
static void infinity()
|
||||
{
|
||||
printf("%s%s %s %s\n%s%s%s%s\n", tab2, "static",
|
||||
declarator<T>::type_name, "infinity() throw()",
|
||||
tab2, "{ return static_cast<", declarator<T>::type_name,
|
||||
">(0); }");
|
||||
printf("%s%s %s %s\n%s%s%s%s\n", tab2, "static",
|
||||
declarator<T>::type_name, "infinity() throw()", tab2,
|
||||
"{ return static_cast<", declarator<T>::type_name, ">(0); }");
|
||||
}
|
||||
|
||||
static void quiet_NaN()
|
||||
{
|
||||
printf("%s%s %s %s\n%s%s%s%s\n", tab2, "static",
|
||||
declarator<T>::type_name, "quiet_NaN() throw()",
|
||||
tab2, "{ return static_cast<", declarator<T>::type_name,
|
||||
">(0); }");
|
||||
printf("%s%s %s %s\n%s%s%s%s\n", tab2, "static",
|
||||
declarator<T>::type_name, "quiet_NaN() throw()",
|
||||
tab2, "{ return static_cast<", declarator<T>::type_name,
|
||||
">(0); }");
|
||||
}
|
||||
|
||||
|
||||
static void signaling_NaN()
|
||||
{
|
||||
printf("%s%s %s %s\n%s%s%s%s\n", tab2, "static",
|
||||
declarator<T>::type_name, "signaling_NaN() throw()",
|
||||
tab2, "{ return static_cast<", declarator<T>::type_name,
|
||||
">(0); }");
|
||||
printf("%s%s %s %s\n%s%s%s%s\n", tab2, "static",
|
||||
declarator<T>::type_name, "signaling_NaN() throw()",
|
||||
tab2, "{ return static_cast<", declarator<T>::type_name,
|
||||
">(0); }");
|
||||
}
|
||||
|
||||
static void denorm_min()
|
||||
{
|
||||
printf("%s%s %s %s\n%s%s%s%s\n", tab2, "static",
|
||||
declarator<T>::type_name, "denorm_min() throw()",
|
||||
tab2, "{ return static_cast<", declarator<T>::type_name,
|
||||
">(0); }");
|
||||
printf("%s%s %s %s\n%s%s%s%s\n", tab2, "static",
|
||||
declarator<T>::type_name, "denorm_min() throw()",
|
||||
tab2, "{ return static_cast<", declarator<T>::type_name,
|
||||
">(0); }");
|
||||
}
|
||||
};
|
||||
};
|
||||
|
||||
|
||||
template<typename T>
|
||||
void is_iec559()
|
||||
{
|
||||
printf("%s%s = %s;\n", tab2, "static const bool is_iec559",
|
||||
bool_alpha[predicate<T>::is_iec559]);
|
||||
printf("%s%s = %s;\n", tab2, "static const bool is_iec559",
|
||||
bool_alpha[predicate<T>::is_iec559]);
|
||||
}
|
||||
|
||||
// tinyness_before
|
||||
template<typename T>
|
||||
void tinyness_before()
|
||||
{
|
||||
printf("%s%s;\n", tab2, "static const bool tinyness_before = false");
|
||||
printf("%s%s;\n", tab2, "static const bool tinyness_before = false");
|
||||
}
|
||||
|
||||
// round style
|
||||
template<typename T>
|
||||
void round_style()
|
||||
{
|
||||
printf("%s%s;\n", tab2, "static const float_round_style "
|
||||
"round_style = round_toward_zero");
|
||||
printf("%s%s;\n", tab2, "static const float_round_style "
|
||||
"round_style = round_toward_zero");
|
||||
}
|
||||
|
||||
|
||||
|
||||
// type traits
|
||||
template<typename T> struct type_trait {
|
||||
|
||||
template<typename T>
|
||||
struct type_trait
|
||||
{
|
||||
type_trait()
|
||||
{
|
||||
declarator<T>::start();
|
||||
printf("%s%s;\n\n", tab2, "static const bool is_specialized = true");
|
||||
min<T>();
|
||||
max<T>();
|
||||
printf("\n");
|
||||
digits<T>();
|
||||
digits10<T>();
|
||||
is_signed<T>();
|
||||
is_integer<T>();
|
||||
is_exact<T>();
|
||||
radix<T>();
|
||||
epsilon<T>();
|
||||
round_error<T>();
|
||||
printf("\n");
|
||||
min_exponent<T>();
|
||||
min_exponent10<T>();
|
||||
max_exponent<T>();
|
||||
max_exponent10<T>();
|
||||
printf("\n");
|
||||
has_infinity<T>();
|
||||
has_quiet_nan<T>();
|
||||
has_signaling_nan<T>();
|
||||
infinity_trait<T>::has_denorm();
|
||||
has_denorm_loss<T>();
|
||||
printf("\n");
|
||||
infinity_trait<T>::infinity();
|
||||
infinity_trait<T>::quiet_NaN();
|
||||
infinity_trait<T>::signaling_NaN();
|
||||
infinity_trait<T>::denorm_min();
|
||||
printf("\n");
|
||||
is_iec559<T>();
|
||||
printf("%s%s;\n", tab2, "static const bool is_bounded = true");
|
||||
is_modulo<T>();
|
||||
printf("\n");
|
||||
traps<T>();
|
||||
tinyness_before<T>();
|
||||
round_style<T>();
|
||||
declarator<T>::end();
|
||||
declarator<T>::start();
|
||||
printf("%s%s;\n\n", tab2, "static const bool is_specialized = true");
|
||||
min<T>();
|
||||
max<T>();
|
||||
printf("\n");
|
||||
digits<T>();
|
||||
digits10<T>();
|
||||
is_signed<T>();
|
||||
is_integer<T>();
|
||||
is_exact<T>();
|
||||
radix<T>();
|
||||
epsilon<T>();
|
||||
round_error<T>();
|
||||
printf("\n");
|
||||
min_exponent<T>();
|
||||
min_exponent10<T>();
|
||||
max_exponent<T>();
|
||||
max_exponent10<T>();
|
||||
printf("\n");
|
||||
has_infinity<T>();
|
||||
has_quiet_nan<T>();
|
||||
has_signaling_nan<T>();
|
||||
infinity_trait<T>::has_denorm();
|
||||
has_denorm_loss<T>();
|
||||
printf("\n");
|
||||
infinity_trait<T>::infinity();
|
||||
infinity_trait<T>::quiet_NaN();
|
||||
infinity_trait<T>::signaling_NaN();
|
||||
infinity_trait<T>::denorm_min();
|
||||
printf("\n");
|
||||
is_iec559<T>();
|
||||
printf("%s%s;\n", tab2, "static const bool is_bounded = true");
|
||||
is_modulo<T>();
|
||||
printf("\n");
|
||||
traps<T>();
|
||||
tinyness_before<T>();
|
||||
round_style<T>();
|
||||
declarator<T>::end();
|
||||
}
|
||||
};
|
||||
};
|
||||
|
||||
int main()
|
||||
{
|
||||
type_trait<bool>();
|
||||
|
||||
type_trait<char>();
|
||||
type_trait<signed char>();
|
||||
type_trait<unsigned char>();
|
||||
type_trait<bool>();
|
||||
|
||||
type_trait<char>();
|
||||
type_trait<signed char>();
|
||||
type_trait<unsigned char>();
|
||||
#if defined( _GLIBCPP_USE_WCHAR_T)
|
||||
type_trait<wchar_t>();
|
||||
type_trait<wchar_t>();
|
||||
#endif
|
||||
|
||||
type_trait<short>();
|
||||
type_trait<unsigned short>();
|
||||
|
||||
type_trait<int>();
|
||||
type_trait<unsigned int>();
|
||||
|
||||
type_trait<long>();
|
||||
type_trait<unsigned long>();
|
||||
|
||||
|
||||
type_trait<short>();
|
||||
type_trait<unsigned short>();
|
||||
|
||||
type_trait<int>();
|
||||
type_trait<unsigned int>();
|
||||
|
||||
type_trait<long>();
|
||||
type_trait<unsigned long>();
|
||||
|
||||
#ifdef _GLIBCPP_USE_LONG_LONG
|
||||
type_trait<long long>();
|
||||
type_trait<unsigned long long>();
|
||||
type_trait<long long>();
|
||||
type_trait<unsigned long long>();
|
||||
#endif
|
||||
|
||||
type_trait<float>();
|
||||
type_trait<double>();
|
||||
type_trait<long double>();
|
||||
|
||||
// x86/linux gets this weirdness for the min/max functions:
|
||||
// static long double min() throw()
|
||||
// { return (__extension__ ((union __convert_long_double)
|
||||
// {__convert_long_double_i: {0x0, 0x80000000, 0x1, 0x0}})
|
||||
// .__convert_long_double_d); }
|
||||
type_trait<float>();
|
||||
type_trait<double>();
|
||||
type_trait<long double>();
|
||||
|
||||
// x86/linux gets this weirdness for the min/max functions:
|
||||
// static long double min() throw()
|
||||
// { return (__extension__ ((union __convert_long_double)
|
||||
// {__convert_long_double_i: {0x0, 0x80000000, 0x1, 0x0}})
|
||||
// .__convert_long_double_d); }
|
||||
}
|
||||
|
||||
// G++ doesn't have support for automatic instantiation of static data
|
||||
|
|
|
|||
Loading…
Reference in New Issue