This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
[C++] fix 20647
- From: Nathan Sidwell <nathan at codesourcery dot com>
- To: gcc-patches at gcc dot gnu dot org
- Date: Sun, 22 Oct 2006 17:09:29 +0100
- Subject: [C++] fix 20647
I've installed this patch to fix bug 20647. The typeinfo type string is not
affected by incompleteness of the type itself.
built & tested on i686-pc-linux-gnu. I'll patch 4.1 and 4.2 presently.
nathan
--
Nathan Sidwell :: http://www.codesourcery.com :: CodeSourcery
nathan@codesourcery.com :: http://www.planetfall.pwp.blueyonder.co.uk
2006-10-22 Nathan Sidwell <nathan@codesourcery.com>
PR c++/20647
* rtti.c (tinfo_base_init): The type info string is always global.
PR c++/20647
* g++.dg/abi/rtti3.C: New.
Index: gcc/cp/rtti.c
===================================================================
--- gcc/cp/rtti.c (revision 117956)
+++ gcc/cp/rtti.c (working copy)
@@ -824,13 +824,7 @@ tinfo_base_init (tinfo_s *ti, tree targe
TREE_STATIC (name_decl) = 1;
DECL_EXTERNAL (name_decl) = 0;
DECL_TINFO_P (name_decl) = 1;
- if (involves_incomplete_p (target))
- {
- TREE_PUBLIC (name_decl) = 0;
- DECL_INTERFACE_KNOWN (name_decl) = 1;
- }
- else
- set_linkage_according_to_type (target, name_decl);
+ set_linkage_according_to_type (target, name_decl);
import_export_decl (name_decl);
DECL_INITIAL (name_decl) = name_string;
mark_used (name_decl);
Index: gcc/testsuite/g++.dg/abi/rtti3.C
===================================================================
--- gcc/testsuite/g++.dg/abi/rtti3.C (revision 0)
+++ gcc/testsuite/g++.dg/abi/rtti3.C (revision 0)
@@ -0,0 +1,13 @@
+// PR 20647, we must emit the typeinfo's string as weak, but not the
+// necessarily the type info object
+
+// { dg-require-weak "" }
+// { dg-final { scan-assembler ".weak\[ \t\]_?_ZTSPP1A" } }
+// { dg-final { scan-assembler-not ".weak\[ \t\]_?_ZTIPP1A" } }
+
+struct A;
+
+void Foo ()
+{
+ throw (A **)0;
+}