Re: long double and GLIBCPP_CHECK_MATH_SUPPORT

[David Edelsohn <dje at watson dot ibm dot com>]

>>>>> Benjamin Kosnik writes:

Ben> although that's definitely the direction i want to go, i'm not quite sure 
Ben> it's the most elegant. In particular, I don't really want to use the 
Ben> mangled macros for each of the long double functions in particular, 
Ben> because it will be a mess. (ie, sinl -> _GLIBCPP_HAVE_SINL in std_cmath.h). 

	I do not completely follow.  std_cmath.h already wraps the long
double functions in BOTH

#ifdef _GLIBCPP_USE_LONG_DOUBLE

and

#if _GLIBCPP_HAVE_SINL

Are you planning to remove the mangling that already is present?

	What I am suggesting is that if std_cmath.h already wraps the
collection of long double functions in the _GLIBCPP_USE_LONG_DOUBLE, what
is the point of testing individually for their existence in configure if
_GLIBCPP_USE_LONG_DOUBLE test in configure already failed?

Thanks, David