|
Boost : |
Subject: Re: [boost] threads: linking on OSX
From: Paul A. Bristow (pbristow_at_[hidden])
Date: 2009-10-05 04:53:56
> -----Original Message-----
> From: boost-bounces_at_[hidden] [mailto:boost-bounces_at_[hidden]] On
Behalf Of
> David M. Cotter
> Sent: Saturday, October 03, 2009 11:34 PM
> To: boost_at_[hidden]
> Subject: Re: [boost] threads: linking on OSX
>
> > bjam ... architecture=combined
>
> okay now i get this:
>
> > darwin.compile.c++ bin.v2/libs/math/config/darwin-4.2.1/release/
> > architecture-combined/link-static/threading-multi/
> > has_long_double_support.o
> > /Users/davec/Developer/depot/kJams/External/boost_1_40_0/libs/math/
> > config/has_long_double_support.cpp:9:2: error: #error "long double
> > support is not supported by Boost.Math on this Plaform: the long
> > double version of the TR1 library will not be built."
> > lipo: can't figure out the architecture type of: /var/folders/h7/
> > h7zL6S-RECGa80fEMYXIpk+++TI/-Tmp-//cc9IHJPJ.out
>
> is that a bad thing?
the config file contains
#ifdef BOOST_MATH_NO_LONG_DOUBLE_MATH_FUNCTIONS
#error "long double support is not supported by Boost.Math on this Plaform: the
long double version of the TR1 library will not be built."
#endif
So the error message means BOOST_MATH_NO_LONG_DOUBLE_MATH_FUNCTIONS must be
defined, so that you won't get long double support.
I suspect this is because you are using Darwin.
I:\boost-sandbox\math_toolkit\boost\math\tools\config.hpp
In it John Maddock notes:
// Darwin's rather strange "double double" is rather hard to
// support, it should be possible given enough effort though...
I believe that the upshot is all is well using double and float, but not long
double. If you don't want long double, no problem.
If you really need more accuracy, you could use NTL (for which support is
provided - see the docs).
HTH
Paul
PS John Maddock may well give a more authoritative reply in due course.
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk