Hello,
we currently reject programs of the form
>
> module m
> integer :: s
> end module m
> subroutine s
> use m, only: x => s, x => s
> end subroutine s
with an error stating that S is the name of the current program unit.
Interestingly, the duplicate rename is necessary to trigger it.
There doesn't seem to be a consensus as to whether it should be
accepted, but I think it should be.
Quoting Dominique's comment in the PR:
> if
>
> use m, only: A => X
> use m, only: B => X
>
> is valid, I don't see why
>
> use m, only: A => X
> use m, only: A => X
>
> should not.
The problem is we check the original (symbol) name instead of the local
(symtree) name.
The fix is close to obvious, and should be safe for the branches (this
is a regression)
Regression tested on x86_64-linux. OK for trunk/4.9/4.8 ?
Mikael
2015-02-06 Mikael Morin <mik...@gcc.gnu.org>
PR fortran/63744
* module.c (check_for_ambiguous): Change argument type
from gfc_symbol to gfc_symtree. Check local (symtree) name
instead of original (symbol) name.
(read_module): Update caller.
2015-02-06 Mikael Morin <mik...@gcc.gnu.org>
PR fortran/63744
gfortran.dg/use_rename_8.f90: New.
Index: module.c
===================================================================
--- module.c (révision 220107)
+++ module.c (copie de travail)
@@ -4795,19 +4795,21 @@ read_cleanup (pointer_info *p)
/* It is not quite enough to check for ambiguity in the symbols by
the loaded symbol and the new symbol not being identical. */
static bool
-check_for_ambiguous (gfc_symbol *st_sym, pointer_info *info)
+check_for_ambiguous (gfc_symtree *st, pointer_info *info)
{
gfc_symbol *rsym;
module_locus locus;
symbol_attribute attr;
+ gfc_symbol *st_sym;
- if (gfc_current_ns->proc_name && st_sym->name == gfc_current_ns->proc_name->name)
+ if (gfc_current_ns->proc_name && st->name == gfc_current_ns->proc_name->name)
{
gfc_error ("%qs of module %qs, imported at %C, is also the name of the "
- "current program unit", st_sym->name, module_name);
+ "current program unit", st->name, module_name);
return true;
}
+ st_sym = st->n.sym;
rsym = info->u.rsym.sym;
if (st_sym == rsym)
return false;
@@ -5037,7 +5039,7 @@ read_module (void)
if (st != NULL)
{
/* Check for ambiguous symbols. */
- if (check_for_ambiguous (st->n.sym, info))
+ if (check_for_ambiguous (st, info))
st->ambiguous = 1;
else
info->u.rsym.symtree = st;
! { dg-do compile }
!
! PR fortran/63744
! duplicate use rename used to be rejected when the target name
! was that of the current program unit
!
! Original testcase from Roger Ferrer Ibanez <roger.fer...@bsc.es>
MODULE MOO
INTEGER :: A, B, C, D, E, F, G, H, I
END MODULE MOO
SUBROUTINE S
USE MOO, ONLY: X => A, X => A
END SUBROUTINE S
SUBROUTINE T
USE MOO, ONLY: X => B
USE MOO, ONLY: X => B
END SUBROUTINE T
SUBROUTINE C
USE MOO, ONLY: C ! { dg-error "is also the name of the current program unit" }
END SUBROUTINE C
SUBROUTINE D
USE MOO, ONLY: X => D
END SUBROUTINE D
SUBROUTINE E
USE MOO, ONLY: X => E, X => E
END SUBROUTINE E
SUBROUTINE F
USE MOO, ONLY: X => F
USE MOO, ONLY: X => F
END SUBROUTINE F
SUBROUTINE X
USE MOO, ONLY: X => G ! { dg-error "is also the name of the current program unit" }
END SUBROUTINE X
SUBROUTINE Y
USE MOO, ONLY: Y => H ! { dg-error "is also the name of the current program unit" }
END SUBROUTINE Y
SUBROUTINE Z
USE MOO, ONLY: Z => I, Z => I ! { dg-error "is also the name of the current program unit" }
END SUBROUTINE Z
