Re: [PATCH] rcu: Fix rcu_torture_read ftrace event

[Paul E. McKenney]

On Mon, Mar 20, 2023 at 01:36:50PM -0400, Steven Rostedt wrote:
> On Mon, 20 Mar 2023 09:58:11 -0700
> "Paul E. McKenney" <paulmck@xxxxxxxxxx> wrote:
> 
> > On Mon, Mar 20, 2023 at 11:20:15AM -0400, Steven Rostedt wrote:
> > > 
> > > [ Wondering why this didn't get picked up in v6.3-rc3, I see that the
> > >   maintainers of RCU were not Cc'd :-( ]
> > > 
> > > This is a bug that will cause unwanted results. I have a patch that will not
> > > let the kernel build when code like this is added.
> > > 
> > >   https://patchwork.kernel.org/project/linux-trace-kernel/patch/20230309221302.642e82d9@xxxxxxxxxxxxxxxxxx/
> > > 
> > > ( The kernel robot even failed when applying the above patch, because it
> > >   caught the code that this patch fixes )
> > > 
> > > On Mon,  6 Mar 2023 12:27:43 +0000
> > > Douglas RAILLARD <douglas.raillard@xxxxxxx> wrote:
> > >   
> > > > From: Douglas Raillard <douglas.raillard@xxxxxxx>
> > > > 
> > > > Fix the rcutorturename field so that its size is correctly reported in
> > > > the text format embedded in trace.dat files. As it stands, it is
> > > > reported as being of size 1:  
> > > 
> > > And that the offsets of the following fields will be incorrect as well.
> > >   
> > > > 
> > > >     field:char rcutorturename[8];   offset:8;       size:1; signed:0;
> > > >   
> > > 
> > > Please add:
> > > 
> > > Cc: stable@xxxxxxxxxxxxxxx
> > > Fixes: 04ae87a52074e ("ftrace: Rework event_create_dir()")  
> > 
> > Thank you, Steve!
> > 
> > With those fixes, and with an ack or better from Steve, I will be happy
> > to pull this in to -rcu.  How urgent is this?  The default destination
> > would be the v6.5 merge window (not the upcoming one, but the one after
> > that), so if you need it sooner, please let me know.
> 
> I would like my patch to get in this release, so if you can get it into
> this release too (before the next merge window) that would be great. This is
> a real bug. User space tooling can not parse this trace event (when it use
> to, so it is a regression), and my patch that prevents other trace events
> from making the same mistake will make this code as is fail the build.
> 
> For this patch:
> 
>  Reviewed-by: Steven Rostedt (Google) <rostedt@xxxxxxxxxxx>

Fair enough, thank you all!

							Thanx, Paul

> -- Steve
> 
> 
> > >   
> > > > Signed-off-by: Douglas Raillard <douglas.raillard@xxxxxxx>
> > > > ---
> > > >  include/trace/events/rcu.h | 2 +-
> > > >  1 file changed, 1 insertion(+), 1 deletion(-)
> > > > 
> > > > diff --git a/include/trace/events/rcu.h b/include/trace/events/rcu.h
> > > > index 90b2fb0292cb..012fa0d171b2 100644
> > > > --- a/include/trace/events/rcu.h
> > > > +++ b/include/trace/events/rcu.h
> > > > @@ -768,7 +768,7 @@ TRACE_EVENT_RCU(rcu_torture_read,
> > > >  	TP_ARGS(rcutorturename, rhp, secs, c_old, c),
> > > >  
> > > >  	TP_STRUCT__entry(
> > > > -		__field(char, rcutorturename[RCUTORTURENAME_LEN])
> > > > +		__array(char, rcutorturename, RCUTORTURENAME_LEN)
> > > >  		__field(struct rcu_head *, rhp)
> > > >  		__field(unsigned long, secs)
> > > >  		__field(unsigned long, c_old)  
> > >   
>