Re: [PATCH v1 6/8] tracing/ftrace: Add might_fault check to syscall probes

[Mathieu Desnoyers]

On 2024-10-04 00:36, Steven Rostedt wrote:
> On Thu,  3 Oct 2024 11:16:36 -0400
> Mathieu Desnoyers <mathieu.desnoyers@xxxxxxxxxxxx> wrote:
>
> > diff --git a/include/trace/trace_events.h b/include/trace/trace_events.h
> > index 0228d9ed94a3..e0d4850b0d77 100644
> > --- a/include/trace/trace_events.h
> > +++ b/include/trace/trace_events.h
> > @@ -446,6 +446,7 @@ __DECLARE_EVENT_CLASS(call, PARAMS(proto), PARAMS(args), PARAMS(tstruct), \
> >   static notrace void							\
> >   trace_event_raw_event_##call(void *__data, proto)			\
> >   {									\
> > +	might_fault();							\
>
> I don't think we want "might_fault()" here, as this is called for every
> tracepoint that is created by the TRACE_EVENT() macro. That means, there's
> going to be plenty of locations this gets called at that do not allow faults.

Here is the full context where this line applies:

#undef DECLARE_EVENT_SYSCALL_CLASS
#define DECLARE_EVENT_SYSCALL_CLASS(call, proto, args, tstruct, assign, print) \
__DECLARE_EVENT_CLASS(call, PARAMS(proto), PARAMS(args), PARAMS(tstruct), \
                      PARAMS(assign), PARAMS(print))                    \
static notrace void                                                     \
trace_event_raw_event_##call(void *__data, proto)                       \
{                                                                       \
        might_fault();                                                  \
        guard(preempt_notrace)();                                       \
        do_trace_event_raw_event_##call(__data, args);                  \
}

Not an issue, since it's only for syscall tracepoints.

Thanks,

Mathieu

> -- Steve
>
>
> >   	guard(preempt_notrace)();					\
> >   	do_trace_event_raw_event_##call(__data, args);			\
> >   }

--
Mathieu Desnoyers
EfficiOS Inc.
https://www.efficios.com