Re: [PATCH] tracing: tprobe-events: Fix to clean up tprobe correctly when module unload

From: Google
Date: Mon Mar 10 2025 - 01:55:10 EST


On Fri, 7 Mar 2025 12:20:00 -0500
Steven Rostedt <rostedt@xxxxxxxxxxx> wrote:

> On Thu, 6 Mar 2025 15:21:44 +0900
> "Masami Hiramatsu (Google)" <mhiramat@xxxxxxxxxx> wrote:
>
> > When unloading module, the tprobe events are not correctly cleaned
> > up. Thus it becomes `fprobe-event` and never be enabled again even
> > if loading the same module again.
> >
> > For example;
> >
> > # cd /sys/kernel/tracing
> > # modprobe trace_events_sample
> > # echo 't:my_tprobe foo_bar' >> dynamic_events
> > # cat dynamic_events
> > t:tracepoints/my_tprobe foo_bar
> > # rmmod trace_events_sample
> > # cat dynamic_events
> > f:tracepoints/my_tprobe foo_bar
> >
> > As you can see, the second time my_tprobe starts with 'f' instead
> > of 't'.
> >
> > This cleans up (unregister) the tprobe events when module is
> > unloaded.
>
> After applying this, I tried it out:
>
> # cd /sys/kernel/tracing
> # modprobe trace_events_sample
> # echo 't:my_tprobe foo_bar a=$arg1' > dynamic_events
> # echo 1 > events/tracepoints/enable
> [wait]
> # echo 0 > events/tracepoints/enable
> # cat trace
> # tracer: nop
> #
> # entries-in-buffer/entries-written: 13/13 #P:8
> #
> # _-----=> irqs-off/BH-disabled
> # / _----=> need-resched
> # | / _---=> hardirq/softirq
> # || / _--=> preempt-depth
> # ||| / _-=> migrate-disable
> # |||| / delay
> # TASK-PID CPU# ||||| TIMESTAMP FUNCTION
> # | | | ||||| | |
> event-sample-1041 [005] ...1. 1957.632897: my_tprobe: (__probestub_foo_bar+0x4/0x20 [trace_events_sample]) a=0xffffffffc006c186
> event-sample-1041 [005] ...1. 1958.656935: my_tprobe: (__probestub_foo_bar+0x4/0x20 [trace_events_sample]) a=0xffffffffc006c186
> event-sample-1041 [005] ...1. 1959.680883: my_tprobe: (__probestub_foo_bar+0x4/0x20 [trace_events_sample]) a=0xffffffffc006c186
> event-sample-1041 [005] ...1. 1960.704855: my_tprobe: (__probestub_foo_bar+0x4/0x20 [trace_events_sample]) a=0xffffffffc006c186
> event-sample-1041 [005] ...1. 1961.728977: my_tprobe: (__probestub_foo_bar+0x4/0x20 [trace_events_sample]) a=0xffffffffc006c186
> event-sample-1041 [005] ...1. 1962.751822: my_tprobe: (__probestub_foo_bar+0x4/0x20 [trace_events_sample]) a=0xffffffffc006c186
> event-sample-1041 [005] ...1. 1963.776834: my_tprobe: (__probestub_foo_bar+0x4/0x20 [trace_events_sample]) a=0xffffffffc006c186
> event-sample-1041 [005] ...1. 1964.800748: my_tprobe: (__probestub_foo_bar+0x4/0x20 [trace_events_sample]) a=0xffffffffc006c186
> event-sample-1041 [005] ...1. 1965.824788: my_tprobe: (__probestub_foo_bar+0x4/0x20 [trace_events_sample]) a=0xffffffffc006c186
> event-sample-1041 [005] ...1. 1966.848789: my_tprobe: (__probestub_foo_bar+0x4/0x20 [trace_events_sample]) a=0xffffffffc006c186
> event-sample-1041 [005] ...1. 1967.872902: my_tprobe: (__probestub_foo_bar+0x4/0x20 [trace_events_sample]) a=0xffffffffc006c186
> event-sample-1041 [005] ...1. 1968.896754: my_tprobe: (__probestub_foo_bar+0x4/0x20 [trace_events_sample]) a=0xffffffffc006c186
> event-sample-1041 [005] ...1. 1969.920737: my_tprobe: (__probestub_foo_bar+0x4/0x20 [trace_events_sample]) a=0xffffffffc006c186
> # rmmod trace_events_sample
> # cat trace
> # tracer: nop
> #
> # entries-in-buffer/entries-written: 13/13 #P:8
> #
> # _-----=> irqs-off/BH-disabled
> # / _----=> need-resched
> # | / _---=> hardirq/softirq
> # || / _--=> preempt-depth
> # ||| / _-=> migrate-disable
> # |||| / delay
> # TASK-PID CPU# ||||| TIMESTAMP FUNCTION
> # | | | ||||| | |
> event-sample-1041 [005] ...1. 1957.632897: my_tprobe: (0xffffffffc00620b4) a=0xffffffffc006c186
> event-sample-1041 [005] ...1. 1958.656935: my_tprobe: (0xffffffffc00620b4) a=0xffffffffc006c186
> event-sample-1041 [005] ...1. 1959.680883: my_tprobe: (0xffffffffc00620b4) a=0xffffffffc006c186
> event-sample-1041 [005] ...1. 1960.704855: my_tprobe: (0xffffffffc00620b4) a=0xffffffffc006c186
> event-sample-1041 [005] ...1. 1961.728977: my_tprobe: (0xffffffffc00620b4) a=0xffffffffc006c186
> event-sample-1041 [005] ...1. 1962.751822: my_tprobe: (0xffffffffc00620b4) a=0xffffffffc006c186
> event-sample-1041 [005] ...1. 1963.776834: my_tprobe: (0xffffffffc00620b4) a=0xffffffffc006c186
> event-sample-1041 [005] ...1. 1964.800748: my_tprobe: (0xffffffffc00620b4) a=0xffffffffc006c186
> event-sample-1041 [005] ...1. 1965.824788: my_tprobe: (0xffffffffc00620b4) a=0xffffffffc006c186
> event-sample-1041 [005] ...1. 1966.848789: my_tprobe: (0xffffffffc00620b4) a=0xffffffffc006c186
> event-sample-1041 [005] ...1. 1967.872902: my_tprobe: (0xffffffffc00620b4) a=0xffffffffc006c186
> event-sample-1041 [005] ...1. 1968.896754: my_tprobe: (0xffffffffc00620b4) a=0xffffffffc006c186
> event-sample-1041 [005] ...1. 1969.920737: my_tprobe: (0xffffffffc00620b4) a=0xffffffffc006c186
>
> # modprobe trace_events_sample
> # echo 1 > events/tracepoints/enable
> # cat trace
> # tracer: nop
> #
> # entries-in-buffer/entries-written: 13/13 #P:8
> #
> # _-----=> irqs-off/BH-disabled
> # / _----=> need-resched
> # | / _---=> hardirq/softirq
> # || / _--=> preempt-depth
> # ||| / _-=> migrate-disable
> # |||| / delay
> # TASK-PID CPU# ||||| TIMESTAMP FUNCTION
> # | | | ||||| | |
> event-sample-1041 [005] ...1. 1957.632897: my_tprobe: (0xffffffffc00620b4) a=0xffffffffc006c186
> event-sample-1041 [005] ...1. 1958.656935: my_tprobe: (0xffffffffc00620b4) a=0xffffffffc006c186
> event-sample-1041 [005] ...1. 1959.680883: my_tprobe: (0xffffffffc00620b4) a=0xffffffffc006c186
> event-sample-1041 [005] ...1. 1960.704855: my_tprobe: (0xffffffffc00620b4) a=0xffffffffc006c186
> event-sample-1041 [005] ...1. 1961.728977: my_tprobe: (0xffffffffc00620b4) a=0xffffffffc006c186
> event-sample-1041 [005] ...1. 1962.751822: my_tprobe: (0xffffffffc00620b4) a=0xffffffffc006c186
> event-sample-1041 [005] ...1. 1963.776834: my_tprobe: (0xffffffffc00620b4) a=0xffffffffc006c186
> event-sample-1041 [005] ...1. 1964.800748: my_tprobe: (0xffffffffc00620b4) a=0xffffffffc006c186
> event-sample-1041 [005] ...1. 1965.824788: my_tprobe: (0xffffffffc00620b4) a=0xffffffffc006c186
> event-sample-1041 [005] ...1. 1966.848789: my_tprobe: (0xffffffffc00620b4) a=0xffffffffc006c186
> event-sample-1041 [005] ...1. 1967.872902: my_tprobe: (0xffffffffc00620b4) a=0xffffffffc006c186
> event-sample-1041 [005] ...1. 1968.896754: my_tprobe: (0xffffffffc00620b4) a=0xffffffffc006c186
> event-sample-1041 [005] ...1. 1969.920737: my_tprobe: (0xffffffffc00620b4) a=0xffffffffc006c186
>
> Tracing doesn't restart.

Ah, I see. __unregister_trace_fprobe() clean up fp.

static void __unregister_trace_fprobe(struct trace_fprobe *tf)
{
if (trace_fprobe_is_registered(tf)) {
unregister_fprobe(&tf->fp);
memset(&tf->fp, 0, sizeof(tf->fp)); <---- this

This will cleanup the callbacks.

Let me update it.

Thanks!

>
> -- Steve
>


--
Masami Hiramatsu (Google) <mhiramat@xxxxxxxxxx>