[Bug fortran/46937] [4.6 Regression] gfortran.dg/pointer_intent_1.f90 FAILs with -fno-inline

[burnus at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=46937

Tobias Burnus <burnus at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
   Last reconfirmed|                            |2010.12.14 14:36:06
                 CC|                            |burnus at gcc dot gnu.org
   Target Milestone|---                         |4.6.0
     Ever Confirmed|0                           |1

--- Comment #1 from Tobias Burnus <burnus at gcc dot gnu.org> 2010-12-14 
14:36:06 UTC ---
Confirmed. Thanks for the report!

The function annotation patch (PR 43665) was seemingly too eager. The problem
is that the object itself ("t") may not be changed - but pointers it contains
may. In that sense, it resembles a bit PR 45586.

Patch:

--- a/gcc/fortran/trans-types.c
+++ b/gcc/fortran/trans-types.c
@@ -2325,7 +2325,11 @@ create_fn_spec (gfc_symbol *sym, tree fntype)
     if (spec_len < sizeof (spec))
       {
        if (!f->sym || f->sym->attr.pointer || f->sym->attr.target
-           || f->sym->attr.external || f->sym->attr.cray_pointer)
+           || f->sym->attr.external || f->sym->attr.cray_pointer
+           || (f->sym->ts.type == BT_DERIVED
+               && f->sym->ts.u.derived->attr.pointer_comp)
+           || (f->sym->ts.type == BT_CLASS
+               && CLASS_DATA (f->sym)->ts.u.derived->attr.pointer_comp))
          spec[spec_len++] = '.';
        else if (f->sym->attr.intent == INTENT_IN)
          spec[spec_len++] = 'r';