Re: Help! - Problems compiling kde4 on amd64 with gcc 4.5.3

[David Brownlee <abs%netbsd.org@localhost>]

On 27 August 2011 17:53, Dave Tyson <dtyson%wirralcavinggroup.org.uk@localhost> 
wrote:
> Just had a go at building kde4 again on NetBSD current dated 25/8/2011 (amd64)
> using pkgsrc checked out at the same time.
>
> NetBSD skip 5.99.55 NetBSD 5.99.55 (GENERIC) #0: Thu Aug 25 17:28:55 UTC 2011
> builds%b6.netbsd.org@localhost:/home/builds/ab/HEAD/amd64/201108251530Z-
> obj/home/builds/ab/HEAD/src/sys/arch/amd64/compile/GENERIC amd64
>
> kdepimlibs4 drops out with a c++ compilation error:
>
> Building CXX object gpgme++/CMakeFiles/gpgmepp.dir/gpgsignkeyeditinteractor.o
> /usr/pkgsrc/misc/kdepimlibs4/work/kdepimlibs-4.5.5/gpgme++/gpgsignkeyeditinteractor.cpp:
> In member function 'virtual unsigned int
> GpgME::GpgSignKeyEditInteractor::nextState(unsigned int, const char*,
> GpgME::Error&) const':
> /usr/pkgsrc/misc/kdepimlibs4/work/kdepimlibs-4.5.5/gpgme++/gpgsignkeyeditinteractor.cpp:247:138:
> error: no matching function for call to
> 'std::map<boost::tuples::tuple<GpgSignKeyEditInteractor_Private::SignKeyState,
> unsigned int, std::basic_string<char> >,
> GpgSignKeyEditInteractor_Private::SignKeyState>::find(std::tuple<GpgSignKeyEditInteractor_Private::SignKeyState,
> unsigned int, std::basic_string<char> >) const'
> /usr/include/g++/bits/stl_map.h:696:7: note: candidates are: std::map<_Key,
> _Tp, _Compare, _Alloc>::iterator std::map<_Key, _Tp, _Compare,
> _Alloc>::find(const key_type&) [with _Key =
> boost::tuples::tuple<GpgSignKeyEditInteractor_Private::SignKeyState, unsigned
> int, std::basic_string<char> >, _Tp =
> GpgSignKeyEditInteractor_Private::SignKeyState, _Compare =
> std::less<boost::tuples::tuple<GpgSignKeyEditInteractor_Private::SignKeyState,
> unsigned int, std::basic_string<char> > >, _Alloc =
> std::allocator<std::pair<const
> boost::tuples::tuple<GpgSignKeyEditInteractor_Private::SignKeyState, unsigned
> int, std::basic_string<char> >,
> GpgSignKeyEditInteractor_Private::SignKeyState> >, std::map<_Key, _Tp,
> _Compare, _Alloc>::iterator = std::_Rb_tree_iterator<std::pair<const
> boost::tuples::tuple<GpgSignKeyEditInteractor_Private::SignKeyState, unsigned
> int, std::basic_string<char> >,
> GpgSignKeyEditInteractor_Private::SignKeyState> >, key_type =
> boost::tuples::tuple<GpgSignKeyEditInteractor_Private::SignKeyState, unsigned
> int, std::basic_string<char> >]
> /usr/include/g++/bits/stl_map.h:711:7: note:                 std::map<_Key,
> _Tp, _Compare, _Alloc>::const_iterator std::map<_Key, _Tp, _Compare,
> _Alloc>::find(const key_type&) const [with _Key =
> boost::tuples::tuple<GpgSignKeyEditInteractor_Private::SignKeyState, unsigned
> int, std::basic_string<char> >, _Tp =
> GpgSignKeyEditInteractor_Private::SignKeyState, _Compare =
> std::less<boost::tuples::tuple<GpgSignKeyEditInteractor_Private::SignKeyState,
> unsigned int, std::basic_string<char> > >, _Alloc =
> std::allocator<std::pair<const
> boost::tuples::tuple<GpgSignKeyEditInteractor_Private::SignKeyState, unsigned
> int, std::basic_string<char> >,
> GpgSignKeyEditInteractor_Private::SignKeyState> >, std::map<_Key, _Tp,
> _Compare, _Alloc>::const_iterator =
> std::_Rb_tree_const_iterator<std::pair<const
> boost::tuples::tuple<GpgSignKeyEditInteractor_Private::SignKeyState, unsigned
> int, std::basic_string<char> >,
> GpgSignKeyEditInteractor_Private::SignKeyState> >, key_type =
> boost::tuples::tuple<GpgSignKeyEditInteractor_Private::SignKeyState, unsigned
> int, std::basic_string<char> >]
> *** Error code 1
>
> Stop.
> make: stopped in /usr/pkgsrc/misc/kdepimlibs4/work/kdepimlibs-4.5.5/_KDE_build
> *** Error code 1
>
> The line in question looks innocuous:
>
>   const TransitionMap::const_iterator it = table.find( make_tuple(
>        static_cast<SignKeyState>( state() ), status, std::string( args ) ) );
>
> and the same code compiles fine under i386 NetBSD 5.1
>
> I am guessing this may be an issue with gcc 4.5.3, but don't have a clue about
> c++.
>
> Has anyone else seen this problem? Should I submit a bug report or is there a
> subtle fix :-)

I'm seeing the same issue - also on amd64 gcc-4.5.3. I wonder if its
the combination of amd64 & gcc-4.5.3. I have a 5.1 amd64 box I can
test build on to eliminate that. security/gpgme was updated on the
14th of august, so it *could* also be related to that. On your i386
5.1 box are you using the latest pkgsrc?