[Bug fortran/59537] No "Automatic array cannot have an initializer", for -finit-real without a SAVE statement present in subroutine

[dominiq at lps dot ens.fr]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=59537

Dominique d'Humieres <dominiq at lps dot ens.fr> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
            Summary|"An automatic object shall  |No "Automatic array cannot
                   |not have the SAVE           |have an initializer", for
                   |attribute." is not          |-finit-real without a SAVE
                   |diagnosed.                  |statement present in
                   |                            |subroutine

--- Comment #5 from Dominique d'Humieres <dominiq at lps dot ens.fr> ---
>  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".

You are right:

5.4.14 SAVE statement
...
1 A SAVE statement with a saved entity list species the SAVE attribute
(5.3.16) for a list of entities. A SAVE statement without a saved entity list
is treated as though it contained the names of all allowed items in the same
scoping unit.

I thought that the SAVE statement applied to ALL local variables, my bad.

> Nonetheless, code that compiles without -finit-real should also compile
> with -finit-real, right?

I disagree: C506 states that automatic object cannot be initialized. What is
wrong is that it depends on the presence of the SAVE statement without a saved
entity list.