[Dune-devel] [Dune] Problems compiling dune-grid 2.2 with alugrid support using gcc 4.8.0 on mac

[Uli Sack]

Hi,
This holds for gcc 4.7.1 as well. The change described below by Andreas 
needs to go to <aluinst>/include/serial/gitter_tetra_top.h lines 1294 
and 2062 - so it's an ALUGrid thing and has nothing to do with the 
Dune-release branch and/or trunk, or not?!

best,Uli

> Hi,
>
> I had this on clang as well it's fixed in trunk, but has not been
> ported to the 2.2 release branch (yet?). To fix it, add a '<<' in
> front of the __FILE__ macro.
> cheers
>  Andreas


>> Dear dune-list,
>>
>> I just tried to compile dune-grid (from the 2.2 release branch) on
>> my new macbook with gcc version 4.8, more precisely:

>> it does not seem to find alugrid and prints out the following
>> warning:
[...]
>> I attach config.log, the relevant lines appear to be

>> In file included from /usr/local/alugrid-1.50/include
>> /alugrid_serial.h:118:0,
>>                 from conftest.cpp:103:
>> /usr/local/alugrid-1.50/include/serial/gitter_tetra_top.h: In member
>> function 'ALUGridSpace::TetraTop<A>::myhface3_t*
>> ALUGridSpace::TetraTop<A>::subface3(int, int)':
>> /usr/local/alugrid-1.50/include/serial/gitter_tetra_top.h:1294:97:
>> error: unable to find string literal operator 'operator"" __FILE__'
>>     cerr << "**FEHLER (FATAL): Falsche Verfeinerungsregel [" <<
>> this->myhface3(i)->getrule() << "] in "__FILE__ << " " << __LINE__
>> << endl ;
[...]
d
-- 
************************************************************************
* Uli Sack                     ** email: usack at math.fu-berlin.de       *
* Freie Universität Berlin     ** phone: + 49 (30) 838 75250           *
* Institut für Mathematik      **                                      *
* Arnimallee 6                 ** -------------------------------------*
* 14195 Berlin, Germany        ** Member of MATHEON (www.matheon.de)   *
************************************************************************