msvcp: Export the various floating point constants.

dlls/msvcp100/msvcp100.spec

@@ -2788,24 +2788,24 @@
 @ stub -arch=win32 ?xsputn@?$basic_streambuf@_WU?$char_traits@_W@std@@@std@@MAE_JPB_W_J@Z
 @ stub -arch=win64 ?xsputn@?$basic_streambuf@_WU?$char_traits@_W@std@@@std@@MEAA_JPEB_W_J@Z
 @ stub _Cosh
-# extern _Denorm
+@ extern _Denorm msvcp90._Denorm
 @ stub _Dnorm
 @ stub _Dscale
 @ stub _Dtest
-# extern _Eps
+@ extern _Eps msvcp90._Eps
 @ stub _Exp
 @ stub _FCosh
-# extern _FDenorm
+@ extern _FDenorm msvcp90._FDenorm
 @ stub _FDnorm
 @ stub _FDscale
 @ stub _FDtest
-# extern _FEps
+@ extern _FEps msvcp90._FEps
 @ stub _FExp
-# extern _FInf
-# extern _FNan
+@ extern _FInf msvcp90._FInf
+@ extern _FNan msvcp90._FNan
 # extern _FRteps
 @ stub _FSinh
-# extern _FSnan
+@ extern _FSnan msvcp90._FSnan
 # extern _FXbig
 @ stub _GetLocaleForCP
 @ stub _Getcoll
@@ -2814,33 +2814,33 @@
 @ stub _Getdateorder
 @ stub _Getwctype
 @ stub _Getwctypes
-# extern _Hugeval
-# extern _Inf
+@ extern _Hugeval msvcp90._Hugeval
+@ extern _Inf msvcp90._Inf
 @ stub _LCosh
-# extern _LDenorm
+@ extern _LDenorm msvcp90._LDenorm
 @ stub _LDscale
 @ stub _LDtest
-# extern _LEps
+@ extern _LEps msvcp90._LEps
 @ stub _LExp
-# extern _LInf
-# extern _LNan
+@ extern _LInf msvcp90._LInf
+@ extern _LNan msvcp90._LNan
 @ stub _LPoly
 # extern _LRteps
 @ stub _LSinh
-# extern _LSnan
+@ extern _LSnan msvcp90._LSnan
 # extern _LXbig
-# extern _LZero
+@ extern _LZero msvcp90._LZero
 @ stub _Mbrtowc
 @ stub _Mtxdst
 @ stub _Mtxinit
 @ stub _Mtxlock
 @ stub _Mtxunlock
-# extern _Nan
+@ extern _Nan msvcp90._Nan
 @ stub _Once
 @ stub _Poly
 # extern _Rteps
 @ stub _Sinh
-# extern _Snan
+@ extern _Snan msvcp90._Snan
 @ stub _Stod
 @ stub _Stodx
 @ stub _Stof

dlls/msvcp60/msvcp60.spec

@@ -4255,49 +4255,49 @@
 @ stub -arch=win32 ?xsputn@?$basic_streambuf@GU?$char_traits@G@std@@@std@@MAEHPBGH@Z
 @ stub -arch=win64 ?xsputn@?$basic_streambuf@GU?$char_traits@G@std@@@std@@MEAA_JPEBG_J@Z
 @ stub _Cosh
-# extern _Denorm
+@ extern _Denorm msvcp90._Denorm
 @ stub _Dnorm
 @ stub _Dscale
 @ stub _Dtest
-# extern _Eps
+@ extern _Eps msvcp90._Eps
 @ stub _Exp
 @ stub _FCosh
-# extern _FDenorm
+@ extern _FDenorm msvcp90._FDenorm
 @ stub _FDnorm
 @ stub _FDscale
 @ stub _FDtest
-# extern _FEps
+@ extern _FEps msvcp90._FEps
 @ stub _FExp
-# extern _FInf
-# extern _FNan
+@ extern _FInf msvcp90._FInf
+@ extern _FNan msvcp90._FNan
 # extern _FRteps
 @ stub _FSinh
-# extern _FSnan
+@ extern _FSnan msvcp90._FSnan
 # extern _FXbig
 @ stub _Getcoll
 @ stub _Getctype
 @ stub _Getcvt
-# extern _Hugeval
-# extern _Inf
+@ extern _Hugeval msvcp90._Hugeval
+@ extern _Inf msvcp90._Inf
 @ stub _LCosh
-# extern _LDenorm
+@ extern _LDenorm msvcp90._LDenorm
 @ stub _LDscale
 @ stub _LDtest
-# extern _LEps
+@ extern _LEps msvcp90._LEps
 @ stub _LExp
-# extern _LInf
-# extern _LNan
+@ extern _LInf msvcp90._LInf
+@ extern _LNan msvcp90._LNan
 @ stub _LPoly
 # extern _LRteps
 @ stub _LSinh
-# extern _LSnan
+@ extern _LSnan msvcp90._LSnan
 # extern _LXbig
 @ stub _Mbrtowc
-# extern _Nan
+@ extern _Nan msvcp90._Nan
 @ stub _Poly
 # extern _Rteps
 @ stub _Sinh
-# extern _Snan
+@ extern _Snan msvcp90._Snan
 @ stub _Stod
 @ stub _Stof
 @ stub _Stold

dlls/msvcp70/msvcp70.spec

@@ -5041,51 +5041,51 @@
 @ stub -arch=win32 ?xsputn@?$basic_streambuf@_WU?$char_traits@_W@std@@@std@@MAEHPB_WH@Z
 @ stub -arch=win64 ?xsputn@?$basic_streambuf@_WU?$char_traits@_W@std@@@std@@MEAA_JPEB_W_J@Z
 @ stub _Cosh
-# extern _Denorm
+@ extern _Denorm msvcp90._Denorm
 @ stub _Dnorm
 @ stub _Dscale
 @ stub _Dtest
-# extern _Eps
+@ extern _Eps msvcp90._Eps
 @ stub _Exp
 @ stub _FCosh
-# extern _FDenorm
+@ extern _FDenorm msvcp90._FDenorm
 @ stub _FDnorm
 @ stub _FDscale
 @ stub _FDtest
-# extern _FEps
+@ extern _FEps msvcp90._FEps
 @ stub _FExp
-# extern _FInf
-# extern _FNan
+@ extern _FInf msvcp90._FInf
+@ extern _FNan msvcp90._FNan
 # extern _FRteps
 @ stub _FSinh
-# extern _FSnan
+@ extern _FSnan msvcp90._FSnan
 # extern _FXbig
 # extern _FZero
 @ stub _Getcoll
 @ stub _Getctype
 @ stub _Getcvt
-# extern _Hugeval
-# extern _Inf
+@ extern _Hugeval msvcp90._Hugeval
+@ extern _Inf msvcp90._Inf
 @ stub _LCosh
-# extern _LDenorm
+@ extern _LDenorm msvcp90._LDenorm
 @ stub _LDscale
 @ stub _LDtest
-# extern _LEps
+@ extern _LEps msvcp90._LEps
 @ stub _LExp
-# extern _LInf
-# extern _LNan
+@ extern _LInf msvcp90._LInf
+@ extern _LNan msvcp90._LNan
 @ stub _LPoly
 # extern _LRteps
 @ stub _LSinh
-# extern _LSnan
+@ extern _LSnan msvcp90._LSnan
 # extern _LXbig
-# extern _LZero
+@ extern _LZero msvcp90._LZero
 @ stub _Mbrtowc
-# extern _Nan
+@ extern _Nan msvcp90._Nan
 @ stub _Poly
 # extern _Rteps
 @ stub _Sinh
-# extern _Snan
+@ extern _Snan msvcp90._Snan
 @ stub _Stod
 @ stub _Stof
 @ stub _Stold

dlls/msvcp71/msvcp71.spec

@@ -5095,24 +5095,24 @@
 @ stub -arch=win64 ?xsputn@?$basic_streambuf@_WU?$char_traits@_W@std@@@std@@MEAA_JPEB_W_J@Z
 @ stub _Atexit
 @ stub _Cosh
-# extern _Denorm
+@ extern _Denorm msvcp90._Denorm
 @ stub _Dnorm
 @ stub _Dscale
 @ stub _Dtest
-# extern _Eps
+@ extern _Eps msvcp90._Eps
 @ stub _Exp
 @ stub _FCosh
-# extern _FDenorm
+@ extern _FDenorm msvcp90._FDenorm
 @ stub _FDnorm
 @ stub _FDscale
 @ stub _FDtest
-# extern _FEps
+@ extern _FEps msvcp90._FEps
 @ stub _FExp
-# extern _FInf
-# extern _FNan
+@ extern _FInf msvcp90._FInf
+@ extern _FNan msvcp90._FNan
 # extern _FRteps
 @ stub _FSinh
-# extern _FSnan
+@ extern _FSnan msvcp90._FSnan
 # extern _FXbig
 # extern _FZero
 @ stub _Getcoll
@@ -5120,28 +5120,28 @@
 @ stub _Getcvt
 @ stub _Getwctype
 @ stub _Getwctypes
-# extern _Hugeval
-# extern _Inf
+@ extern _Hugeval msvcp90._Hugeval
+@ extern _Inf msvcp90._Inf
 @ stub _LCosh
-# extern _LDenorm
+@ extern _LDenorm msvcp90._LDenorm
 @ stub _LDscale
 @ stub _LDtest
-# extern _LEps
+@ extern _LEps msvcp90._LEps
 @ stub _LExp
-# extern _LInf
-# extern _LNan
+@ extern _LInf msvcp90._LInf
+@ extern _LNan msvcp90._LNan
 @ stub _LPoly
 # extern _LRteps
 @ stub _LSinh
-# extern _LSnan
+@ extern _LSnan msvcp90._LSnan
 # extern _LXbig
-# extern _LZero
+@ extern _LZero msvcp90._LZero
 @ stub _Mbrtowc
-# extern _Nan
+@ extern _Nan msvcp90._Nan
 @ stub _Poly
 # extern _Rteps
 @ stub _Sinh
-# extern _Snan
+@ extern _Snan msvcp90._Snan
 @ stub _Stod
 @ stub _Stof
 @ stub _Stold

dlls/msvcp80/msvcp80.spec

@@ -5701,28 +5701,28 @@
 @ stub -arch=win32 ?xsputn@?$basic_streambuf@_WU?$char_traits@_W@std@@@std@@MAEHPB_WH@Z
 @ stub -arch=win64 ?xsputn@?$basic_streambuf@_WU?$char_traits@_W@std@@@std@@MEAA_JPEB_W_J@Z
 @ stub _Cosh
-# extern _Denorm
+@ extern _Denorm msvcp90._Denorm
 @ stub _Dnorm
 @ stub _Dscale
 @ stub _Dtentox
 @ stub _Dtest
 @ stub _Dunscale
-# extern _Eps
+@ extern _Eps msvcp90._Eps
 @ stub _Exp
 @ stub _FCosh
-# extern _FDenorm
+@ extern _FDenorm msvcp90._FDenorm
 @ stub _FDnorm
 @ stub _FDscale
 @ stub _FDtentox
 @ stub _FDtest
 @ stub _FDunscale
-# extern _FEps
+@ extern _FEps msvcp90._FEps
 @ stub _FExp
-# extern _FInf
-# extern _FNan
+@ extern _FInf msvcp90._FInf
+@ extern _FNan msvcp90._FNan
 # extern _FRteps
 @ stub _FSinh
-# extern _FSnan
+@ extern _FSnan msvcp90._FSnan
 # extern _FXbig
 @ stub _Getcoll
 @ stub _Getctype
@@ -5730,35 +5730,35 @@
 @ stub _Getdateorder
 @ stub _Getwctype
 @ stub _Getwctypes
-# extern _Hugeval
-# extern _Inf
+@ extern _Hugeval msvcp90._Hugeval
+@ extern _Inf msvcp90._Inf
 @ stub _LCosh
-# extern _LDenorm
+@ extern _LDenorm msvcp90._LDenorm
 @ stub _LDscale
 @ stub _LDtentox
 @ stub _LDtest
 @ stub _LDunscale
-# extern _LEps
+@ extern _LEps msvcp90._LEps
 @ stub _LExp
-# extern _LInf
-# extern _LNan
+@ extern _LInf msvcp90._LInf
+@ extern _LNan msvcp90._LNan
 @ stub _LPoly
 # extern _LRteps
 @ stub _LSinh
-# extern _LSnan
+@ extern _LSnan msvcp90._LSnan
 # extern _LXbig
-# extern _LZero
+@ extern _LZero msvcp90._LZero
 @ stub _Mbrtowc
 @ stub _Mtxdst
 @ stub _Mtxinit
 @ stub _Mtxlock
 @ stub _Mtxunlock
-# extern _Nan
+@ extern _Nan msvcp90._Nan
 @ stub _Once
 @ stub _Poly
 # extern _Rteps
 @ stub _Sinh
-# extern _Snan
+@ extern _Snan msvcp90._Snan
 @ stub _Stod
 @ stub _Stodx
 @ stub _Stof

dlls/msvcp90/math.c

@@ -48,6 +48,26 @@ enum std_float_round_style
     round_toward_neg_infinity = 3
 };
 
+/* these are defined as integers but the bit patterns are then interpreted as floats/doubles */
+const DWORD     _FDenorm = 1;
+const ULONGLONG _Denorm  = 1;
+const ULONGLONG _LDenorm = 1;
+const DWORD     _FInf    = 0x7f800000;
+const ULONGLONG _Inf     = (ULONGLONG)0x7ff00000 << 32;
+const ULONGLONG _LInf    = (ULONGLONG)0x7ff00000 << 32;
+const DWORD     _FNan    = 0x7fc00000;
+const ULONGLONG _Nan     = (ULONGLONG)0x7ff80000 << 32;
+const ULONGLONG _LNan    = (ULONGLONG)0x7ff80000 << 32;
+const DWORD     _FSnan   = 0x7f800001;
+const ULONGLONG _Snan    = ((ULONGLONG)0x7ff00000 << 32) + 1;
+const ULONGLONG _LSnan   = ((ULONGLONG)0x7ff00000 << 32) + 1;
+const ULONGLONG _LZero   = 0;
+const ULONGLONG _Hugeval = (ULONGLONG)0x7ff00000 << 32;
+
+const float   _FEps = FLT_EPSILON;
+const double  _Eps  = DBL_EPSILON;
+const LDOUBLE _LEps = LDBL_EPSILON;
+
 /* ?digits10@?$numeric_limits@C@std@@2HB -> public: static int const std::numeric_limits<signed char>::digits10 */
 const int std_numeric_limits_signed_char_digits10 = 2;
 
@@ -426,13 +446,13 @@ LONG __cdecl std_numeric_limits_long_denorm_min(void) { return 0; }
 ULONG __cdecl std_numeric_limits_unsigned_long_denorm_min(void) { return 0; }
 
 /* ?denorm_min@?$numeric_limits@M@std@@SAMXZ -> public: static float __cdecl std::numeric_limits<float>::denorm_min(void) */
-float __cdecl std_numeric_limits_float_denorm_min(void) { static const DWORD dmin = 1; return *(float *)&dmin; }
+float __cdecl std_numeric_limits_float_denorm_min(void) { return *(float *)&_FDenorm; }
 
 /* ?denorm_min@?$numeric_limits@N@std@@SANXZ -> public: static double __cdecl std::numeric_limits<double>::denorm_min(void) */
-double __cdecl std_numeric_limits_double_denorm_min(void) { static const ULONGLONG dmin = 1; return *(double *)&dmin; }
+double __cdecl std_numeric_limits_double_denorm_min(void) { return *(double *)&_Denorm; }
 
 /* ?denorm_min@?$numeric_limits@O@std@@SAOXZ -> public: static long double __cdecl std::numeric_limits<long double>::denorm_min(void) */
-LDOUBLE __cdecl std_numeric_limits_long_double_denorm_min(void) { static const ULONGLONG dmin = 1; return *(LDOUBLE *)&dmin; }
+LDOUBLE __cdecl std_numeric_limits_long_double_denorm_min(void) { return *(LDOUBLE *)&_LDenorm; }
 
 /* ?denorm_min@?$numeric_limits@_J@std@@SA_JXZ -> public: static __int64 __cdecl std::numeric_limits<__int64>::denorm_min(void) */
 __int64 __cdecl std_numeric_limits_int64_denorm_min(void) { return 0; }
@@ -474,13 +494,13 @@ LONG __cdecl std_numeric_limits_long_epsilon(void) { return 0; }
 ULONG __cdecl std_numeric_limits_unsigned_long_epsilon(void) { return 0; }
 
 /* ?epsilon@?$numeric_limits@M@std@@SAMXZ -> public: static float __cdecl std::numeric_limits<float>::epsilon(void) */
-float __cdecl std_numeric_limits_float_epsilon(void) { return FLT_EPSILON; }
+float __cdecl std_numeric_limits_float_epsilon(void) { return _FEps; }
 
 /* ?epsilon@?$numeric_limits@N@std@@SANXZ -> public: static double __cdecl std::numeric_limits<double>::epsilon(void) */
-double __cdecl std_numeric_limits_double_epsilon(void) { return DBL_EPSILON; }
+double __cdecl std_numeric_limits_double_epsilon(void) { return _Eps; }
 
 /* ?epsilon@?$numeric_limits@O@std@@SAOXZ -> public: static long double __cdecl std::numeric_limits<long double>::epsilon(void) */
-LDOUBLE __cdecl std_numeric_limits_long_double_epsilon(void) { return LDBL_EPSILON; }
+LDOUBLE __cdecl std_numeric_limits_long_double_epsilon(void) { return _LEps; }
 
 /* ?epsilon@?$numeric_limits@_J@std@@SA_JXZ -> public: static __int64 __cdecl std::numeric_limits<__int64>::epsilon(void) */
 __int64 __cdecl std_numeric_limits_int64_epsilon(void) { return 0; }
@@ -522,13 +542,13 @@ LONG __cdecl std_numeric_limits_long_infinity(void) { return 0; }
 ULONG __cdecl std_numeric_limits_unsigned_long_infinity(void) { return 0; }
 
 /* ?infinity@?$numeric_limits@M@std@@SAMXZ -> public: static float __cdecl std::numeric_limits<float>::infinity(void) */
-float __cdecl std_numeric_limits_float_infinity(void) { static const DWORD inf = 0x7f800000; return *(float *)&inf; }
+float __cdecl std_numeric_limits_float_infinity(void) { return *(float *)&_FInf; }
 
 /* ?infinity@?$numeric_limits@N@std@@SANXZ -> public: static double __cdecl std::numeric_limits<double>::infinity(void) */
-double __cdecl std_numeric_limits_double_infinity(void) { static const ULONGLONG inf = (ULONGLONG)0x7ff00000 << 32; return *(double *)&inf; }
+double __cdecl std_numeric_limits_double_infinity(void) { return *(double *)&_Inf; }
 
 /* ?infinity@?$numeric_limits@O@std@@SAOXZ -> public: static long double __cdecl std::numeric_limits<long double>::infinity(void) */
-LDOUBLE __cdecl std_numeric_limits_long_double_infinity(void) { static const ULONGLONG inf = (ULONGLONG)0x7ff00000 << 32; return *(LDOUBLE *)&inf; }
+LDOUBLE __cdecl std_numeric_limits_long_double_infinity(void) { return *(LDOUBLE *)&_LInf; }
 
 /* ?infinity@?$numeric_limits@_J@std@@SA_JXZ -> public: static __int64 __cdecl std::numeric_limits<__int64>::infinity(void) */
 __int64 __cdecl std_numeric_limits_int64_infinity(void) { return 0; }
@@ -666,13 +686,13 @@ LONG __cdecl std_numeric_limits_long_quiet_NaN(void) { return 0; }
 ULONG __cdecl std_numeric_limits_unsigned_long_quiet_NaN(void) { return 0; }
 
 /* ?quiet_NaN@?$numeric_limits@M@std@@SAMXZ -> public: static float __cdecl std::numeric_limits<float>::quiet_NaN(void) */
-float __cdecl std_numeric_limits_float_quiet_NaN(void) { static const DWORD qnan = 0x7fc00000; return *(float *)&qnan; }
+float __cdecl std_numeric_limits_float_quiet_NaN(void) { return *(float *)&_FNan; }
 
 /* ?quiet_NaN@?$numeric_limits@N@std@@SANXZ -> public: static double __cdecl std::numeric_limits<double>::quiet_NaN(void) */
-double __cdecl std_numeric_limits_double_quiet_NaN(void) { static const ULONGLONG qnan = (ULONGLONG)0x7ff80000 << 32; return *(double *)&qnan; }
+double __cdecl std_numeric_limits_double_quiet_NaN(void) { return *(double *)&_Nan; }
 
 /* ?quiet_NaN@?$numeric_limits@O@std@@SAOXZ -> public: static long double __cdecl std::numeric_limits<long double>::quiet_NaN(void) */
-LDOUBLE __cdecl std_numeric_limits_long_double_quiet_NaN(void) { static const ULONGLONG qnan = (ULONGLONG)0x7ff80000 << 32; return *(LDOUBLE *)&qnan; }
+LDOUBLE __cdecl std_numeric_limits_long_double_quiet_NaN(void) { return *(LDOUBLE *)&_LNan; }
 
 /* ?quiet_NaN@?$numeric_limits@_J@std@@SA_JXZ -> public: static __int64 __cdecl std::numeric_limits<__int64>::quiet_NaN(void) */
 __int64 __cdecl std_numeric_limits_int64_quiet_NaN(void) { return 0; }
@@ -762,13 +782,13 @@ LONG __cdecl std_numeric_limits_long_signaling_NaN(void) { return 0; }
 ULONG __cdecl std_numeric_limits_unsigned_long_signaling_NaN(void) { return 0; }
 
 /* ?signaling_NaN@?$numeric_limits@M@std@@SAMXZ -> public: static float __cdecl std::numeric_limits<float>::signaling_NaN(void) */
-float __cdecl std_numeric_limits_float_signaling_NaN(void) { static const DWORD snan = 0x7f800001; return *(float *)&snan; }
+float __cdecl std_numeric_limits_float_signaling_NaN(void) { return *(float *)&_FSnan; }
 
 /* ?signaling_NaN@?$numeric_limits@N@std@@SANXZ -> public: static double __cdecl std::numeric_limits<double>::signaling_NaN(void) */
-double __cdecl std_numeric_limits_double_signaling_NaN(void) { static const ULONGLONG snan = ((ULONGLONG)0x7ff00000 << 32) + 1; return *(double *)&snan; }
+double __cdecl std_numeric_limits_double_signaling_NaN(void) { return *(double *)&_Snan; }
 
 /* ?signaling_NaN@?$numeric_limits@O@std@@SAOXZ -> public: static long double __cdecl std::numeric_limits<long double>::signaling_NaN(void) */
-LDOUBLE __cdecl std_numeric_limits_long_double_signaling_NaN(void) { static const ULONGLONG snan = ((ULONGLONG)0x7ff00000 << 32) + 1; return *(LDOUBLE *)&snan; }
+LDOUBLE __cdecl std_numeric_limits_long_double_signaling_NaN(void) { return *(LDOUBLE *)&_LSnan; }
 
 /* ?signaling_NaN@?$numeric_limits@_J@std@@SA_JXZ -> public: static __int64 __cdecl std::numeric_limits<__int64>::signaling_NaN(void) */
 __int64 __cdecl std_numeric_limits_int64_signaling_NaN(void) { return 0; }

dlls/msvcp90/msvcp90.spec

@@ -5731,24 +5731,24 @@
 @ stub -arch=win32 ?xsputn@?$basic_streambuf@_WU?$char_traits@_W@std@@@std@@MAEHPB_WH@Z
 @ stub -arch=win64 ?xsputn@?$basic_streambuf@_WU?$char_traits@_W@std@@@std@@MEAA_JPEB_W_J@Z
 @ stub _Cosh
-# extern _Denorm
+@ extern _Denorm
 @ stub _Dnorm
 @ stub _Dscale
 @ stub _Dtest
-# extern _Eps
+@ extern _Eps
 @ stub _Exp
 @ stub _FCosh
-# extern _FDenorm
+@ extern _FDenorm
 @ stub _FDnorm
 @ stub _FDscale
 @ stub _FDtest
-# extern _FEps
+@ extern _FEps
 @ stub _FExp
-# extern _FInf
-# extern _FNan
+@ extern _FInf
+@ extern _FNan
 # extern _FRteps
 @ stub _FSinh
-# extern _FSnan
+@ extern _FSnan
 # extern _FXbig
 @ stub _Getcoll
 @ stub _Getctype
@@ -5756,33 +5756,33 @@
 @ stub _Getdateorder
 @ stub _Getwctype
 @ stub _Getwctypes
-# extern _Hugeval
-# extern _Inf
+@ extern _Hugeval
+@ extern _Inf
 @ stub _LCosh
-# extern _LDenorm
+@ extern _LDenorm
 @ stub _LDscale
 @ stub _LDtest
-# extern _LEps
+@ extern _LEps
 @ stub _LExp
-# extern _LInf
-# extern _LNan
+@ extern _LInf
+@ extern _LNan
 @ stub _LPoly
 # extern _LRteps
 @ stub _LSinh
-# extern _LSnan
+@ extern _LSnan
 # extern _LXbig
-# extern _LZero
+@ extern _LZero
 @ stub _Mbrtowc
 @ stub _Mtxdst
 @ stub _Mtxinit
 @ stub _Mtxlock
 @ stub _Mtxunlock
-# extern _Nan
+@ extern _Nan
 @ stub _Once
 @ stub _Poly
 # extern _Rteps
 @ stub _Sinh
-# extern _Snan
+@ extern _Snan
 @ stub _Stod
 @ stub _Stodx
 @ stub _Stof