Re: [PATCH] tracing: Avoid adding duplicated tracer options when update_tracer_options is running in parallel

[Steven Rostedt]

On Wed, 23 Mar 2022 22:21:29 +0800
Mark-PK Tsai <mark-pk.tsai@xxxxxxxxxxxx> wrote:

> > On Wed, 23 Mar 2022 11:24:42 +0800
> > Mark-PK Tsai <mark-pk.tsai@xxxxxxxxxxxx> wrote:
> >   
> > > When update_tracer_options is running in parallel,
> > > tr->tops might be updated before the trace_types list traversal.
> > > Let update_tracer_options traverse the trace_types list safely in
> > > kernel init time and avoid the tr->tops update before it finish.  
> > 
> > ??? Have you seen a bug here? I'm totally confused by this.  
> 
> Sorry to make you confused.
> 
> After the below patch, update_tracer_options might be executed later than registering
> hwlat_tracer, which is in late_initcall.
> 
> https://lore.kernel.org/lkml/20220316151639.9216-1-mark-pk.tsai@xxxxxxxxxxxx/

If you send patches that depend on patches that are not in the tree, you
need to explicitly state that.


> 
> The init_hwlat_tracer initcall will put hwlat_tracer to tr->tops.
> Then when the later arrived __update_tracer_options is trying to
> update all the tracer options, create_trace_option_files show the
> below warning because hwlat_tracer is already in the list.
> 
> [ 6.680068 ][ T7 ] WARNING: CPU: 0 PID: 7 at kernel/trace/trace.c:8899 create_trace_option_files (kernel/trace/trace.c:8899 (discriminator 1))
> 
> full log: https://lore.kernel.org/lkml/20220322133339.GA32582@xsang-OptiPlex-9020/

So this is all dependent on patches not in the tree?

> 
> 
> >   
> > > 
> > > Link: https://lore.kernel.org/lkml/20220322133339.GA32582@xsang-OptiPlex-9020/
> > > Reported-by: kernel test robot <oliver.sang@xxxxxxxxx>
> > > Signed-off-by: Mark-PK Tsai <mark-pk.tsai@xxxxxxxxxxxx>
> > > ---
> > >  kernel/trace/trace.c | 7 +++++++
> > >  1 file changed, 7 insertions(+)
> > > 
> > > diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c
> > > index adb37e437a05..2974ae056068 100644
> > > --- a/kernel/trace/trace.c
> > > +++ b/kernel/trace/trace.c
> > > @@ -6317,12 +6317,18 @@ static void tracing_set_nop(struct trace_array *tr)
> > >  	tr->current_trace = &nop_trace;
> > >  }
> > >  
> > > +static bool tracer_options_updated;
> > > +
> > >  static void add_tracer_options(struct trace_array *tr, struct tracer *t)
> > >  {
> > >  	/* Only enable if the directory has been created already. */
> > >  	if (!tr->dir)
> > >  		return;
> > >  
> > > +	/* Only create trace option files after update_tracer_options finish */
> > > +	if (!tracer_options_updated)
> > > +		return;
> > > +
> > >  	create_trace_option_files(tr, t);
> > >  }
> > >  
> > > @@ -9133,6 +9139,7 @@ static void update_tracer_options(struct trace_array *tr)
> > >  {
> > >  	mutex_lock(&trace_types_lock);  
> > 
> > How is update_trace_options run in parallel?
> > 
> > There's a mutex that protects it. 
> >   
> 
> Oh sorry.
> What I trying to tell is that update_trace_options is run in parallel with
> the initcall thread after:
> 
> https://lore.kernel.org/lkml/20220316151639.9216-1-mark-pk.tsai@xxxxxxxxxxxx/
> 

Again, this is not in the tree, so it should be part of that patch series,
which I haven't yet been able to fully review.

-- Steve