I think you can use FTRACE (http://lwn.net/Articles/365835/) and set function_graph as the tracer. To filter the output, you must enable CONFIG_DYNAMIC_FTRACE and set which function should be traced. On Sun, Mar 29, 2015 at 5:16 PM, Andy Ng <andreas2025@xxxxxxxxx> wrote: > Hello, > > I am using an RT kernel and in my driver there is an atomic code > path that just makes some critical I/O operations. I would like to > trigger a system function trace when kernel enters into a specific > function in my driver and stop the system trace at the end of the > function. I would like to identify time execution latencies for a > specific code path. I am using Cortex A9 and I have perf and PMU up > and running, but still it is not clear to me how to retrieve the > traces with those latencies in my code path. > > Any ideas how to trace/measure that will be very much appreciated. > > Best regards, > Andy > -- > To unsubscribe from this list: send the line "unsubscribe linux-rt-users" in > the body of a message to majordomo@xxxxxxxxxxxxxxx > More majordomo info at http://vger.kernel.org/majordomo-info.html -- To unsubscribe from this list: send the line "unsubscribe linux-rt-users" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
