On Wed, 2012-03-28 at 06:00 -0400, Luis R. Rodriguez wrote:

I've managed to backport tracing it seems, even when CONFIG_TRACEPOINTS
is enabled.

Nope, you've managed to completely kill tracing.

+#undef __DECLARE_TRACE
+#define __DECLARE_TRACE(name, proto, args, cond, data_proto, data_args) \
+	static inline void trace_##name(proto)				\
+	{ }								\
+	static inline void trace_##name##_rcuidle(proto)		\
+	{ }								\
+	static inline int						\
+	register_trace_##name(void (*probe)(proto))			\
+	{								\
+		return -ENOSYS;						\
+	}								\
+	static inline void						\
+	unregister_trace_##name(void (*probe)(proto))			\
+	{								\
+		return -ENOSYS;						\
+	}								\
+	static inline void check_trace_callback_type_##name(void (*cb)(data_proto)) \
+	{								\
+	}

All of this simply completely kills tracing. Worse, the tracepoints get
registered, but their trace_##name functions do nothing, so you can
think that they're there, but they never record data ...

The old way was much better. Can we just let 2.6.24 and the other
ancients die in peace?

johannes