This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: [Patch, Fortran] SELECT TYPE with associate-name
On 10/06/2009 04:52 PM, Daniel Kraft wrote:
> Is the patch ok in the attached form? (I have regtested it again
> successfully.)
>
> I think it's once again ok, but please also give Tobias another chance
> to look at your patch from his point of view. I did not in detail
> study your solutions to the problems he found with your first patch.
Also OK from my side if you include the original test case (which you
did not attach this time) and with a test case for this BLOCK problem.
Some item which may need to be solved later is:
C812 (R822) If selector is not a variable or is a variable that has a
vector subscript, associate-name shall not appear in a variable
definition context (16.5.7).
I have not tried it but I think the current implementation allows to
violate this constraint.
Tobias
PS: I still do not like
+ /* Special case: If we're in a SELECT TYPE block,
+ replace the selector variable by a temporary. */
+ if (gfc_current_state () == COMP_SELECT_TYPE
+ && st && st->n.sym == type_selector)
+ st = select_type_tmp;
but that block of code is only moved and a separate issue.