[PATCH v2] sched/tracing: Show stacktrace for wakeup tracers

[PATCH v2] sched/tracing: Show stacktrace for wakeup tracers

[Changbin Du]

This align the behavior of wakeup tracers with irqsoff latency tracer
that we record stacktrace at the beginning and end of waking up. The
stacktrace shows us what is happening in the kernel.

Signed-off-by: Changbin Du <changbin.du@gmail.com>
---
 kernel/trace/trace_sched_wakeup.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/kernel/trace/trace_sched_wakeup.c b/kernel/trace/trace_sched_wakeup.c
index da5b6e012840..f4fe7d1781e9 100644
--- a/kernel/trace/trace_sched_wakeup.c
+++ b/kernel/trace/trace_sched_wakeup.c
@@ -475,6 +475,7 @@ probe_wakeup_sched_switch(void *ignore, bool preempt,
 
 	__trace_function(wakeup_trace, CALLER_ADDR0, CALLER_ADDR1, flags, pc);
 	tracing_sched_switch_trace(wakeup_trace, prev, next, flags, pc);
+	__trace_stack(wakeup_trace, flags, 0, pc);
 
 	T0 = data->preempt_timestamp;
 	T1 = ftrace_now(cpu);
@@ -586,6 +587,7 @@ probe_wakeup(void *ignore, struct task_struct *p)
 	data = per_cpu_ptr(wakeup_trace->trace_buffer.data, wakeup_cpu);
 	data->preempt_timestamp = ftrace_now(cpu);
 	tracing_sched_wakeup_trace(wakeup_trace, p, current, flags, pc);
+	__trace_stack(wakeup_trace, flags, 0, pc);
 
 	/*
 	 * We must be careful in using CALLER_ADDR2. But since wake_up
-- 
2.17.1

Re: [PATCH v2] sched/tracing: Show stacktrace for wakeup tracers

[Steven Rostedt]

On Thu, 17 Jan 2019 00:02:49 +0800
Changbin Du <changbin.du@gmail.com> wrote:

> This align the behavior of wakeup tracers with irqsoff latency tracer
> that we record stacktrace at the beginning and end of waking up. The
> stacktrace shows us what is happening in the kernel.
> 
> Signed-off-by: Changbin Du <changbin.du@gmail.com>

I've applied your patch, but it may be a while before you see it in
linux-next. I'll be traveling for a bit and wont be able to do the full
tests which I do before pushing to my linux-next branch.

You may see them soon in my ftrace/core branch on kernel.org.

-- Steve

> ---
>  kernel/trace/trace_sched_wakeup.c | 2 ++
>  1 file changed, 2 insertions(+)
> 
> diff --git a/kernel/trace/trace_sched_wakeup.c b/kernel/trace/trace_sched_wakeup.c
> index da5b6e012840..f4fe7d1781e9 100644
> --- a/kernel/trace/trace_sched_wakeup.c
> +++ b/kernel/trace/trace_sched_wakeup.c
> @@ -475,6 +475,7 @@ probe_wakeup_sched_switch(void *ignore, bool preempt,
>  
>  	__trace_function(wakeup_trace, CALLER_ADDR0, CALLER_ADDR1, flags, pc);
>  	tracing_sched_switch_trace(wakeup_trace, prev, next, flags, pc);
> +	__trace_stack(wakeup_trace, flags, 0, pc);
>  
>  	T0 = data->preempt_timestamp;
>  	T1 = ftrace_now(cpu);
> @@ -586,6 +587,7 @@ probe_wakeup(void *ignore, struct task_struct *p)
>  	data = per_cpu_ptr(wakeup_trace->trace_buffer.data, wakeup_cpu);
>  	data->preempt_timestamp = ftrace_now(cpu);
>  	tracing_sched_wakeup_trace(wakeup_trace, p, current, flags, pc);
> +	__trace_stack(wakeup_trace, flags, 0, pc);
>  
>  	/*
>  	 * We must be careful in using CALLER_ADDR2. But since wake_up