|
Boost Users : |
From: Eoin (eoin-keyword-boostusers.07781a_at_[hidden])
Date: 2006-09-05 17:08:34
Hello I am developing a library which uses a variant like this one
class my_class {};
typedef boost::make_recursive_variant<
my_class,
int ,
std::string,
boost::posix_time::ptime,
std::vector<boost::recursive_variant_>,
std::map<std::string, boost::recursive_variant_>
>::type param;
Under a Debug build when the type my_class is not present everything
compiles perfectly, however with my_class I get internal compiler errors
such as the ones below;
G:\Program Files (x86)\Microsoft Visual Studio .NET
2003\Vc7\include\xstring(827) : fatal error C1001: INTERNAL COMPILER ERROR
(compiler file 'msc1.cpp', line 2701)
Please choose the Technical Support command on the Visual C++
Help menu, or open the Technical Support help file for more
information
c1xx : fatal error C1063: INTERNAL COMPILER ERROR
Please choose the Technical Support command on the Visual C++
Help menu, or open the Technical Support help file for more
information
While playing around with the code and making adjustments (like
reordering the types for the variant) I once got the compiler to display
this error;
fatal error C1067: compiler limit : debug information module size exceeded
Searchs on the net pointed to using the #pragma component (mintypeinfo,
on ) around the offending code, but even when I include this pragam
enclosing the entire header containing my variant type I continue to get
the Internal Compiler Error for some files, though it does go away for
others.
As hinted to by searches on the net this problem seems to be related to
debug information and indeed everything does work fine under a release
build. However as I am developing a library for use by others I cannot
force them to avoid debug builds :^) . Everything works fine with the
lastest MinGW debug or release.
Any suggestions would be very much appreciated. Kind Regards, Eoin.
Boost-users list run by williamkempf at hotmail.com, kalb at libertysoft.com, bjorn.karlsson at readsoft.com, gregod at cs.rpi.edu, wekempf at cox.net