This is the mail archive of the
gdb-patches@sourceware.org
mailing list for the GDB project.
Re: FYI: another fix for PR 12533
>>>>> "Ulrich" == Ulrich Weigand <uweigand@de.ibm.com> writes:
Ulrich> I'm now seeing build failures on RHEL5:
[...]
Sorry about that.
Ulrich> I guess res_val still ought to be NULL initialized here, in the
Ulrich> (rather theoretical)
Ulrich> case the switch falls into an unhandled value ...
I agree.
I'm checking in the appended. I verified that it fixes the problem.
Tom
2012-01-04 Tom Tromey <tromey@redhat.com>
* python/py-value.c (valpy_binop): Initialize 'res_val'.
Index: python/py-value.c
===================================================================
RCS file: /cvs/src/src/gdb/python/py-value.c,v
retrieving revision 1.32
diff -u -r1.32 py-value.c
--- python/py-value.c 3 Jan 2012 19:27:52 -0000 1.32
+++ python/py-value.c 4 Jan 2012 16:13:59 -0000
@@ -725,7 +725,7 @@
{
struct value *arg1, *arg2;
struct cleanup *cleanup = make_cleanup_value_free_to_mark (value_mark ());
- struct value *res_val;
+ struct value *res_val = NULL;
/* If the gdb.Value object is the second operand, then it will be passed
to us as the OTHER argument, and SELF will be an entirely different