On Apr 5, 2010, at 11:50 AM, Bill Hoffman wrote: > Alan W. Irwin wrote: > >>> Have you run cmake --trace yet to figure out what is including what and >>> where? >> Yes, as alluded to above >> the complete result for cmake --trace --debug-output for >> a CMakeLists.txt file consisting of just >> project(test NONE) >> enable_language(CXX) >> was attached to my first post in this thread. Part way through is starts >> accessing CMakeFortranInformation.cmake which is the source of the trouble. >> For your convenience I will attach that cmake output to this post as well. > > OK, you are going to have to look at the C++ code to figure this out. > > You will need to add some debug prints to this function: > > void > cmGlobalGenerator::EnableLanguage(std::vector<std::string>const& languages, > cmMakefile *mf, bool) > > ... > > for(std::vector<std::string>::const_iterator l = languages.begin(); > l != languages.end(); ++l) > { > > Seems that Fortran is getting into the list of enabled languages. > > Might also want to put prints in cmEnableLanguageCommand > ::InitialPass and void cmMakefile::EnableLanguage(std::vector<std::string> > const & lang, > bool optional) > >
So interleaving wine tracing and cmake tracing along with some debugging... the problem comes from within SystemTools::GetActualCaseForPath() where it takes as input <somepath>/CMakeTestCXXCompiler.cmake and returns <somepath>/CMakeFortranInformation.cmake. I'm guessing its a bug in Wine's GetShortPathName() or GetLongPathName(). There are more modern apis for getting the actual case of a file, but they probably don't work on Windows 98. Clint _______________________________________________ Powered by www.kitware.com Visit other Kitware open-source projects at http://www.kitware.com/opensource/opensource.html Please keep messages on-topic and check the CMake FAQ at: http://www.cmake.org/Wiki/CMake_FAQ Follow this link to subscribe/unsubscribe: http://www.cmake.org/mailman/listinfo/cmake