tracing/filters: clean up filter_add_subsystem_pred()

Impact: cleanup, memory leak fix

This patch cleans up filter_add_subsystem_pred():

- searches for the field before creating a copy of the pred

- fixes memory leak in the case a predicate isn't applied

- if -ENOMEM, makes sure there's no longer a reference to the
  pred so the caller can free the half-finished filter

- changes the confusing i == MAX_FILTER_PRED - 1 comparison
  previously remarked upon

This affects only per-subsystem event filtering.

Signed-off-by: Tom Zanussi <tzanussi@gmail.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: =?ISO-8859-1?Q?Fr=E9d=E9ric?= Weisbecker <fweisbec@gmail.com>
LKML-Reference: <1237796808.7527.40.camel@charm-linux>
Signed-off-by: Ingo Molnar <mingo@elte.hu>

kernel/trace/trace_events.c

@@ -585,6 +585,7 @@ subsystem_filter_write(struct file *filp, const char __user *ubuf, size_t cnt,
 	}
 
 	if (filter_add_subsystem_pred(system, pred)) {
+		filter_free_subsystem_preds(system);
 		filter_free_pred(pred);
 		return -EINVAL;
 	}