[PATCH] tracing: Increase stack skip on event trigger

[PATCH] tracing: Increase stack skip on event trigger

[Namhyung Kim]

When 'stacktrace' trigger is used for event tracing, it should skip 4
functions instead of 3.  Otherwise it'll contain ftrace_raw_event_XXX
like below:

  # echo stacktrace > events/sched/sched_switch/trigger
  # tail trace
   => rcu_gp_kthread
   => kthread
   => ret_from_fork
       ksoftirqd/0-3     [000] d..3    53.343993: <stack trace>
   => ftrace_raw_event_sched_switch
   => __schedule
   => schedule
   => smpboot_thread_fn
   => kthread
   => ret_from_fork

Reported-by: Minchan Kim <minchan@kernel.org>
Cc: Tom Zanussi <tom.zanussi@linux.intel.com>
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
---
 kernel/trace/trace_events_trigger.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/kernel/trace/trace_events_trigger.c b/kernel/trace/trace_events_trigger.c
index 8712df9decb4..c58fbdd3053e 100644
--- a/kernel/trace/trace_events_trigger.c
+++ b/kernel/trace/trace_events_trigger.c
@@ -965,12 +965,13 @@ static __init int register_trigger_snapshot_cmd(void) { return 0; }
 
 #ifdef CONFIG_STACKTRACE
 /*
- * Skip 3:
+ * Skip 4:
  *   stacktrace_trigger()
  *   event_triggers_post_call()
+ *   ftrace_event_buffer_commit()
  *   ftrace_raw_event_xxx()
  */
-#define STACK_SKIP 3
+#define STACK_SKIP 4
 
 static void
 stacktrace_trigger(struct event_trigger_data *data)
-- 
2.3.3

Re: [PATCH] tracing: Increase stack skip on event trigger

[Steven Rostedt]

On Thu, 26 Mar 2015 09:39:23 +0900
Namhyung Kim <namhyung@kernel.org> wrote:

> When 'stacktrace' trigger is used for event tracing, it should skip 4
> functions instead of 3.  Otherwise it'll contain ftrace_raw_event_XXX
> like below:
> 
>   # echo stacktrace > events/sched/sched_switch/trigger
>   # tail trace
>    => rcu_gp_kthread
>    => kthread
>    => ret_from_fork
>        ksoftirqd/0-3     [000] d..3    53.343993: <stack trace>
>    => ftrace_raw_event_sched_switch
>    => __schedule
>    => schedule
>    => smpboot_thread_fn
>    => kthread
>    => ret_from_fork
> 
> Reported-by: Minchan Kim <minchan@kernel.org>
> Cc: Tom Zanussi <tom.zanussi@linux.intel.com>
> Signed-off-by: Namhyung Kim <namhyung@kernel.org>
> ---

Hmm, this breaks when frame pointers are not enabled.

I may come up with something though.

-- Steve