http://gcc.gnu.org/bugzilla/show_bug.cgi?id=59537
--- Comment #4 from Lorenz Hüdepohl <bugs at stellardeath dot org> --- > From the above gfortran is right to reject the code with initialization. > > > Without the SAVE statement it compiles fine. > > From the above (C513) this is a bug, the code should give an error > without/with -finit-*. I guess my point is, that I do _not_ specify real, save :: a(1:n) which would be illegal, but merely real :: a(1:n) save in which case the "SAVE" statement should apply to all variables to which it can apply - that is _not_ to the automatic dummy argument "a". Nonetheless, code that compiles without -finit-real should also compile with -finit-real, right?
