http://gcc.gnu.org/bugzilla/show_bug.cgi?id=59939
--- Comment #9 from Zhendong Su <su at cs dot ucdavis.edu> --- (In reply to Marc Glisse from comment #8) > I think Andrew's point is that we might have: > ... Thanks for your explanation Marc. I think I see the point now, although the rational still feels a bit convoluted. > As with many things, it is a matter of taste whether you want to ignore > warnings in obviously dead code. Yes, I very much agree --- many times it's purely a design decision. But please allow me to argue the following point. Since "fn1(a,b)" is obviously dead code in the originally reported test (because of which the two conversions weren't warned), the dead code should then be warned, correct?
