------- Comment #6 from jv244 at cam dot ac dot uk 2009-08-27 16:01 -------
this is bold guess at a patch, which does fix the lineno info for this
testcase, but no idea if this is even remotely correct. Based on the
observation that gfc_match_function_decl sets declared_at, but
gfc_match_subroutine does not.
Index: decl.c
===================================================================
--- decl.c (revision 151059)
+++ decl.c (working copy)
@@ -5047,7 +5047,10 @@
match is_bind_c;
char peek_char;
bool allow_binding_name;
+ locus old_loc;
+ old_loc = gfc_current_locus;
+
if (gfc_current_state () != COMP_NONE
&& gfc_current_state () != COMP_INTERFACE
&& gfc_current_state () != COMP_CONTAINS)
@@ -5147,6 +5150,8 @@
/* Warn if it has the same name as an intrinsic. */
warn_intrinsic_shadow (sym, false);
+ sym->declared_at=old_loc;
+
return MATCH_YES;
}
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=40823
