https://gcc.gnu.org/bugzilla/show_bug.cgi?id=60637
--- Comment #6 from Jonathan Wakely <redi at gcc dot gnu.org> --- (In reply to Marc Glisse from comment #5) > Why do you want to use a different code for C++98 and C++11? The best > solution would be to remove the declaration in C++98, since it is not > standard, but if we have to keep it, Yes, we can't really remove it now, we have no idea how much code is using it. > it would seem easier to share the code > with C++11, no? Maybe. Since I don't plan to change trunk (because the c++98 signbit function template works correctly there) I wanted to make a less obtrusive change than replacing the existing function with the c++11 versions. By just fixing the body we don't change what is declared, we just make it work correctly. Just sharing the code with C++11 might not be straightforward, because I don't know if Solaris 12 declares std::signbit for C++98 mode. If it doesn't, then we can't just reuse the same code, because that code relies on declarations in Solaris' <math.h> That might not be a problem, if _GLIBCXX_USE_C99_MATH is not declared for Solaris 12, because we wouldn't declare std::signbit at all in c++98 mode ... but I don't know, and would rather not take risks or require new configure checks on the stable 4.9 and 5 release branches.
