http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50810
--- Comment #18 from Manuel López-Ibáñez <manu at gcc dot gnu.org> 2011-10-24 12:35:49 UTC --- Paolo, I am sorry to say it so bluntly but this solution is nonsense: http://gcc.gnu.org/ml/gcc-patches/2011-10/msg02160.html I know it is what Gabriel requested, but it doesn't make sense. * It doesn't make sense to have a warning conditional of an option when the negated option cannot disable it. Just don't give it any option then. * It doesn't make sense that the warning is emitted with option Wnarrowing but it is disabled by option -Wno-c++0x-compat. This also breaks consistency with the pragmas and with -Werror=c++0x-compat versus -Werror=narrowing. * It doesn't make sense to force users to see a warning when they have requested to not see it using -Wno-narrowing! Honestly, listen to Jason's stated opinion here: http://gcc.gnu.org/bugzilla/show_bug.cgi?id=49793#c2 This PR is only about adding -Wnarrowing to -Wc++0x-compat to get a warning when not using -std=c++0x. Nothing else. It was fixed by your first patch (plus not enabling -Wc++0x-compat with -Wall). If you have asked Jason, that patch would have been pre-approved and committed long ago!
