On Sat, 03 Dec 2022 12:20:15 +0100
<gregkh@xxxxxxxxxxxxxxxxxxx> wrote:
> The patch below does not apply to the 4.19-stable tree.
> If someone wants it applied there, or to any other stable or longterm
> tree, then please email the backport, including the original git commit
> id to <stable@xxxxxxxxxxxxxxx>.
>
> Possible dependencies:
>
> 4313e5a61304 ("tracing: Free buffers when a used dynamic event is removed")
Hmm, isn't the above the patch that failed to apply?
> 5448d44c3855 ("tracing: Add unified dynamic event framework")
And this is mentioned below.
[..]
> If any dynamic event that is being removed was enabled, then make sure the
> buffers they were enabled in are now cleared.
>
> Link: https://lkml.kernel.org/r/20221123171434.545706e3@xxxxxxxxxxxxxxxxxx
> Link: https://lore.kernel.org/all/20221110020319.1259291-1-zhengyejian1@xxxxxxxxxx/
>
> Cc: stable@xxxxxxxxxxxxxxx
> Cc: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
> Depends-on: e18eb8783ec49 ("tracing: Add tracing_reset_all_online_cpus_unlocked() function")
> Depends-on: 5448d44c38557 ("tracing: Add unified dynamic event framework")
^^^
-- Steve
> Depends-on: 6212dd29683ee ("tracing/kprobes: Use dyn_event framework for kprobe events")
> Depends-on: 065e63f951432 ("tracing: Only have rmmod clear buffers that its events were active in")
> Depends-on: 575380da8b469 ("tracing: Only clear trace buffer on module unload if event was traced")
> Fixes: 77b44d1b7c283 ("tracing/kprobes: Rename Kprobe-tracer to kprobe-event")
> Reported-by: Zheng Yejian <zhengyejian1@xxxxxxxxxx>
> Reported-by: Yujie Liu <yujie.liu@xxxxxxxxx>
> Reported-by: kernel test robot <yujie.liu@xxxxxxxxx>
> Acked-by: Masami Hiramatsu (Google) <mhiramat@xxxxxxxxxx>
> Signed-off-by: Steven Rostedt (Google) <rostedt@xxxxxxxxxxx>
>
> diff --git a/kernel/trace/trace_dynevent.c b/kernel/trace/trace_dynevent.c
> index 154996684fb5..4376887e0d8a 100644
> --- a/kernel/trace/trace_dynevent.c
> +++ b/kernel/trace/trace_dynevent.c
> @@ -118,6 +118,7 @@ int dyn_event_release(const char *raw_command, struct dyn_event_operations *type
> if (ret)
> break;
> }
> + tracing_reset_all_online_cpus();
> mutex_unlock(&event_mutex);
> out:
> argv_free(argv);
> @@ -214,6 +215,7 @@ int dyn_events_release_all(struct dyn_event_operations *type)
> break;
> }
> out:
> + tracing_reset_all_online_cpus();
> mutex_unlock(&event_mutex);
>
> return ret;
> diff --git a/kernel/trace/trace_events.c b/kernel/trace/trace_events.c
> index 78cd19e31dba..f71ea6e79b3c 100644
> --- a/kernel/trace/trace_events.c
> +++ b/kernel/trace/trace_events.c
> @@ -2880,7 +2880,10 @@ static int probe_remove_event_call(struct trace_event_call *call)
> * TRACE_REG_UNREGISTER.
> */
> if (file->flags & EVENT_FILE_FL_ENABLED)
> - return -EBUSY;
> + goto busy;
> +
> + if (file->flags & EVENT_FILE_FL_WAS_ENABLED)
> + tr->clear_trace = true;
> /*
> * The do_for_each_event_file_safe() is
> * a double loop. After finding the call for this
> @@ -2893,6 +2896,12 @@ static int probe_remove_event_call(struct trace_event_call *call)
> __trace_remove_event_call(call);
>
> return 0;
> + busy:
> + /* No need to clear the trace now */
> + list_for_each_entry(tr, &ftrace_trace_arrays, list) {
> + tr->clear_trace = false;
> + }
> + return -EBUSY;
> }
>
> /* Remove an event_call */
