[PATCH RESEND] trace: function graph: Fix filters for function_graph threshold

From: Joel Fernandes
Date: Sat Jun 18 2016 - 01:45:29 EST


Function graph tracer currently ignores filters if tracing_thresh is set.
For example, even if set_ftrace_pid is set, then its ignored if tracing_thresh
set, resulting in all processes being traced.

To fix this, we reuse the same entry function as when tracing_thresh is not
set and do everything as in the regular case except for writing the function entry
to the ring buffer.

Cc: Frederic Weisbecker <fweisbec@xxxxxxxxx>
Cc: Steven Rostedt <rostedt@xxxxxxxxxxx>
Cc: Ingo Molnar <mingo@xxxxxxxxxx>
Signed-off-by: Joel Fernandes <agnel.joel@xxxxxxxxx>
---
kernel/trace/trace_functions_graph.c | 17 ++++++++---------
1 file changed, 8 insertions(+), 9 deletions(-)

diff --git a/kernel/trace/trace_functions_graph.c b/kernel/trace/trace_functions_graph.c
index a663cbb..0967fac 100644
--- a/kernel/trace/trace_functions_graph.c
+++ b/kernel/trace/trace_functions_graph.c
@@ -337,6 +337,13 @@ int trace_graph_entry(struct ftrace_graph_ent *trace)
if (ftrace_graph_notrace_addr(trace->func))
return 1;

+ /*
+ * Stop here if tracing_threshold is set. We only write function return
+ * events to the ring buffer.
+ */
+ if (tracing_thresh)
+ return 1;
+
local_irq_save(flags);
cpu = raw_smp_processor_id();
data = per_cpu_ptr(tr->trace_buffer.data, cpu);
@@ -354,14 +361,6 @@ int trace_graph_entry(struct ftrace_graph_ent *trace)
return ret;
}

-static int trace_graph_thresh_entry(struct ftrace_graph_ent *trace)
-{
- if (tracing_thresh)
- return 1;
- else
- return trace_graph_entry(trace);
-}
-
static void
__trace_graph_function(struct trace_array *tr,
unsigned long ip, unsigned long flags, int pc)
@@ -456,7 +455,7 @@ static int graph_trace_init(struct trace_array *tr)
set_graph_array(tr);
if (tracing_thresh)
ret = register_ftrace_graph(&trace_graph_thresh_return,
- &trace_graph_thresh_entry);
+ &trace_graph_entry);
else
ret = register_ftrace_graph(&trace_graph_return,
&trace_graph_entry);
--
2.5.0