[Bug fortran/56261] [OOP] seg fault call procedure pointer on polymorphic array

[janus at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=56261

--- Comment #9 from janus at gcc dot gnu.org 2013-02-09 16:58:54 UTC ---
(In reply to comment #8)
> On the test case in comment 2,

comment 1?


> ifort v11.1 reports:
> 
> > ifort -o bug.exe bug.F90 
> bug.F90(23): error #6592: This symbol must be a defined parameter, an
> enumerator, or an argument of an inquiry function that evaluates to a
> compile-time constant.   [FF]
>   procedure(  ), pointer                   :: f => ff
> ---------------------------------------------------^
> bug.F90(23): error #6973: This is not a valid initialization expression.   [FF]
>   procedure(  ), pointer                   :: f => ff
> ---------------------------------------------------^
> compilation aborted for bug.F90 (code 1)
> 
> Same for the scalar case.

Thanks for checking. Probably ifort does not support pointer initialization yet
(which is a F08 feature).


> Interestingly, the workaround doesn't work under ifort - it seems not to like:
> 
>   procedure(ff), pointer :: f => ff
> 
> but instead needs:
> 
>   procedure(ff), pointer :: f
>   f => ff

Again, pointer initialization.


> In fact, if I use:
> 
>   procedure(), pointer :: f
>   f => ff
> 
> then it compiles without any warnings/errors but segfaults at runtime.
> 
> With:
> 
> 
>   procedure(ff), pointer :: f
>   f => ff
> 
> it compiles and runs as expected.

Good, that's compatible with gfortran's behavior (which is fine, since the test
case is invalid). Apparently ifort also lacks diagnostics for this.