This is the mail archive of the
libstdc++@sourceware.cygnus.com
mailing list for the libstdc++ project.
Re: <limits> and Cygwin
- To: Scott Ladd <scottrobertladd@yahoo.com>
- Subject: Re: <limits> and Cygwin
- From: Benjamin Kosnik <bkoz@cygnus.com>
- Date: Tue, 31 Aug 1999 16:34:02 -0700 (PDT)
- cc: libstdc++@sourceware.cygnus.com
Gaby's the person with the definitive say on this, but it looks like the
generated specializations for floating types are incorrect.
This is also a problem under a linux native build: The
generated std_limits.h file has these values, which seem off if you
take a look at the example for floats in section 18.2.1.5 of the
standard:
static const bool has_infinity = false;
static const bool has_quiet_NaN = false;
static const bool has_signaling_NaN = false;
static const bool has_denorm_loss = false;
static const bool is_iec559 = false;
static const bool traps = false;
static const bool tinyness_before = false;
Gaby?
> Using gcc 2.95 for Cygwin under Win NT 4, I cannot compile code
> including <limits>. The problem lies in std_limits.h, in the definition
> of numeric_limits<long double>.min, where it complains about a union:
Odd. Can you post the generated specialization for numeric_limits<long
double> from your std_limits.h file?
-Benjamin