This is the mail archive of the gcc-bugs@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

[Bug fortran/51788] Unstable gfc_verify_binding_labels check for a binding label which collides with a global entity


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51788

--- Comment #1 from Tobias Burnus <burnus at gcc dot gnu.org> 2012-01-07 23:15:58 UTC ---
The problems is that:
            if ((sym->attr.use_assoc && bind_c_sym->mod_name
                 && strcmp (bind_c_sym->mod_name, sym->module) != 0)
is not true as the gsym and the sym are for the same symbol.

However, one has an ambiguity:
  (gdb) p sym->ns->sym_root->right->right->name
  $44 = 0x2aaaacc36018 "one"
  (gdb) p sym->ns->sym_root->right->right->ambiguous
  $45 = 1

This ambiguity will later lead to:
  Error: Name 'one' at (1) is an ambiguous reference to 'one'
if one tries to reference the variable.


I do not see a good solution but as the ambiguity is diagnosed if one tries to
reference the variable, I am tempted to use the following patch:


--- a/gcc/testsuite/gfortran.dg/binding_label_tests_10_main.f03
+++ b/gcc/testsuite/gfortran.dg/binding_label_tests_10_main.f03
@@ -7 +7 @@ module binding_label_tests_10_main
-  integer(c_int), bind(c,name="c_one") :: one ! { dg-error "collides" }
+  integer(c_int), bind(c,name="c_one") :: two ! { dg-error "collides" }

--- a/gcc/testsuite/gfortran.dg/binding_label_tests_11_main.f03
+++ b/gcc/testsuite/gfortran.dg/binding_label_tests_11_main.f03
@@ -8,2 +8,2 @@ contains
-  function one() bind(c, name="c_one") ! { dg-error "collides" }
-    integer(c_int) one
+  function two() bind(c, name="c_one") ! { dg-error "collides" }
+    integer(c_int) two
@@ -11 +11 @@ contains
-  end function one
+  end function two


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]