You can forget about this problem. I was able to replace __gnu_cxx::hash_map with boost::unordered_map and compile cleanly. Anyways, it would be a good idea to move to std::tr1 because the ext/hash_xxx headers are already deprecated in gcc 4.3 and are likely to be dropped in a near future. Given that boost::unordered_map compiles just renaming some types and namespaces it shouldn't be hard.
Cheers, Juan Juan Hernando escribió: > Dear all, > I want to debug a random segfault that I'm having inside a third party > library and I want to get the a compilation with the highest number of > runtime debug checks possible. The reason is that valgrind has failed to > diagnose my problem, the crash is random and the backtrace not very > helpful, so I'm a bit lost. > > I've already compiled the other C++ libraries with _GLIBCXX_DEBUG to add > the STL debug code, but I didn't manage to do it with Equalizer. After > some naïve tries I've decided to ask for some advice. > > I'm using the 0.6 stable release and g++ 4.2.3 in ubuntu. The problem is > related to the hash_map extension. I attach the compiler command line > and output. > Does anybody have any suggestion about how to solve the problem or > workaround it? I don't mind lacking debug support for the hash_table > because my bug is not there, but I need a consistent build. > > Of course, this only happens with _GLIBCXX_DEBUG, and that makes me > think there's something awkward inside stdlibc++ but I'd prefer not > messing there. > > Thanks in advance, > Juan > > > ------------------------------------------------------------------------ > > _______________________________________________ > eq-dev mailing list > [email protected] > http://www.equalizergraphics.com/cgi-bin/mailman/listinfo/eq-dev > http://www.equalizergraphics.com _______________________________________________ eq-dev mailing list [email protected] http://www.equalizergraphics.com/cgi-bin/mailman/listinfo/eq-dev http://www.equalizergraphics.com
