Re: [boost] Re: mpl::if_ and ICE triggered on GNU g++-cvs-today

From: Martin Wille (mw8329_at_[hidden])
Date: 2003-07-10 11:04:43

Markus Werle wrote:

>
  There is something wrong with the config.
> Obviously the code should use the
> 
> BOOST_MPL_AUX_VALUE_WKND(C)::value
> 
> but it seems the output of my configure run is not 
> included.
...
> 
> result of configure run (user.hpp):
...
> //
> // options added by configure:
> //
> #define BOOST_MSVC6_MEMBER_TEMPLATES
> #define BOOST_HAS_UNISTD_H
> #define BOOST_HAS_STDINT_H
> #define BOOST_HAS_SIGACTION
> #define BOOST_HAS_SCHED_YIELD
> #define BOOST_HAS_PTHREADS
> #define BOOST_HAS_PTHREAD_MUTEXATTR_SETTYPE
> #define BOOST_HAS_PARTIAL_STD_ALLOCATOR
> #define BOOST_HAS_NRVO
> #define BOOST_HAS_NL_TYPES_H
> #define BOOST_HAS_NANOSLEEP
> #define BOOST_HAS_LONG_LONG
> #define BOOST_HAS_GETTIMEOFDAY
> #define BOOST_HAS_DIRENT_H
> #define BOOST_HAS_CLOCK_GETTIME
> #define BOOST_NO_SWPRINTF
> #define BOOST_NO_STDC_NAMESPACE
> #define BOOST_NO_STD_WSTRING
> #define BOOST_NO_STD_WSTREAMBUF
> #define BOOST_NO_STD_ALLOCATOR
> #define BOOST_NO_CWCTYPE

That is the output for gcc-3.4 from today? I've run the configure
script with gcc version 3.4 20030710 (ie. from today) and got these
results:

#define BOOST_MSVC6_MEMBER_TEMPLATES
#define BOOST_HAS_UNISTD_H
#define BOOST_HAS_STDINT_H
#define BOOST_HAS_SIGACTION
#define BOOST_HAS_SCHED_YIELD
#define BOOST_HAS_PTHREADS
#define BOOST_HAS_PTHREAD_YIELD
#define BOOST_HAS_PTHREAD_MUTEXATTR_SETTYPE
#define BOOST_HAS_PARTIAL_STD_ALLOCATOR
#define BOOST_HAS_NRVO
#define BOOST_HAS_NL_TYPES_H
#define BOOST_HAS_NANOSLEEP
#define BOOST_HAS_LONG_LONG
#define BOOST_HAS_GETTIMEOFDAY
#define BOOST_HAS_DIRENT_H
#define BOOST_HAS_CLOCK_GETTIME
#define BOOST_NO_STD_ALLOCATOR

The BOOST_NO_STD_ALLOCATOR seems to get defined due to a compiler
bug. Otherwise, that output looks ok to me. (The output is the
same for Boost-1.30 and for the current CVS version of Boost.)

Regards,
m