------- Comment #3 from burnus at gcc dot gnu dot org 2008-10-16 09:41 -------
Hmm, maybe the wording could (should?) be changed to make clear that for
subroutine sub(n)
integer n
real automatic_data_object(n)
end subroutine sub
the automatic_data_object gets no SAVE with -fno-automatic. (It doesn't and I
have no idea how it could, but seemingly one can read the option such that it
would.)-- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=37835
