This is the mail archive of the
gdb-patches@sourceware.org
mailing list for the GDB project.
[patch] Fix modification of cplus_struct_default
- From: Jan Kratochvil <jan dot kratochvil at redhat dot com>
- To: gdb-patches at sourceware dot org
- Date: Sun, 27 Jun 2010 15:17:47 +0200
- Subject: [patch] Fix modification of cplus_struct_default
Hi,
(gdb) set print object on
(gdb) print rtti_data
$2 = (RTTI_base) {_vptr.RTTI_base = 0x400bd0}
->
(gdb) set print object on
(gdb) print rtti_data
$2 = {base = {_vptr.RTTI_base = 0x400bd0}, data = 1}
as rtti_data has no RTTI and it falsely interpreted RTTI from its first data
field. gnuv3_dynamic_class has protection against !TYPE_CPLUS_DYNAMIC - but
it could not work.
No regressions on {x86_64,x86_64-m32,i686}-fedora13-linux-gnu.
Thanks,
Jan
gdb/
2010-06-27 Jan Kratochvil <jan.kratochvil@redhat.com>
Fix modification of cplus_struct_default.
* dwarf2read.c (dwarf2_add_member_fn) <no DW_AT_vtable_elem_location>:
Call ALLOCATE_CPLUS_STRUCT_TYPE.
* gdbtypes.c (cplus_struct_default): New empty initializer, comment it.
gdb/testsuite/
2010-06-27 Jan Kratochvil <jan.kratochvil@redhat.com>
* gdb.cp/virtbase.cc (class RTTI_base, class RTTI_data)
(main) <rtti_data>: New.
* gdb.cp/virtbase.exp (print rtti_data): New.
--- a/gdb/dwarf2read.c
+++ b/gdb/dwarf2read.c
@@ -4954,6 +4954,7 @@ dwarf2_add_member_fn (struct field_info *fip, struct die_info *die,
complaint (&symfile_complaints,
_("Member function \"%s\" (offset %d) is virtual but the vtable offset is not specified"),
fieldname, die->offset);
+ ALLOCATE_CPLUS_STRUCT_TYPE (type);
TYPE_CPLUS_DYNAMIC (type) = 1;
}
}
--- a/gdb/gdbtypes.c
+++ b/gdb/gdbtypes.c
@@ -1733,7 +1733,8 @@ check_stub_method_group (struct type *type, int method_id)
}
}
-const struct cplus_struct_type cplus_struct_default;
+/* Ensure it is in .rodata (if available) by workarounding GCC PR 44690. */
+const struct cplus_struct_type cplus_struct_default = { };
void
allocate_cplus_struct_type (struct type *type)
--- a/gdb/testsuite/gdb.cp/virtbase.cc
+++ b/gdb/testsuite/gdb.cp/virtbase.cc
@@ -74,8 +74,19 @@ public:
virtual void b() {}
};
+class RTTI_base
+{
+public:
+ virtual ~RTTI_base() {}
+};
-
+class RTTI_data
+{
+public:
+ RTTI_base base;
+ int data;
+ RTTI_data() : data(1) {}
+};
int main() {
ph::Derived tst;
@@ -84,6 +95,7 @@ int main() {
E *e = new E;
RHB *b = new RHC();
+ RTTI_data rtti_data;
return 0; // breakpoint 3
}
--- a/gdb/testsuite/gdb.cp/virtbase.exp
+++ b/gdb/testsuite/gdb.cp/virtbase.exp
@@ -60,3 +60,8 @@ gdb_test "print *(D *) e" " = {<C> = {v = 11}, _vptr.D = $hex}"
# https://bugzilla.redhat.com/show_bug.cgi?id=560741
gdb_test_no_output "set print object on"
gdb_test "print/x b->mA" " = 0xaaaaaaaa"
+
+# A regression test reported to Red Hat bugzilla, see:
+# https://bugzilla.redhat.com/show_bug.cgi?id=606660
+# `set print object on' is expected.
+gdb_test "print rtti_data" " = .*, data = 1\}"