Re: [PATCH v3 2/8] tracing/ftrace: guard syscall probe with preempt_notrace

[Mathieu Desnoyers]

On 2024-10-09 01:19, Steven Rostedt wrote:
> On Fri,  4 Oct 2024 10:58:12 -0400
> Mathieu Desnoyers <mathieu.desnoyers@xxxxxxxxxxxx> wrote:
>
> > +		      PARAMS(assign), PARAMS(print))			\
> > +static notrace void							\
> > +trace_event_raw_event_##call(void *__data, proto)			\
> > +{									\
> > +	guard(preempt_notrace)();					\
> > +	do_trace_event_raw_event_##call(__data, args);			\
> > +}
> > +
>
> Do we really need to use "guard()" for a single line function? Why make the
> compiler do more work?
>
> static notrace void							\
> trace_event_raw_event_##call(void *__data, proto)			\
> {									\
> 	preempt_disable_notrace();					\
> 	do_trace_event_raw_event_##call(__data, args);			\
> 	preempt_enable_notrace();					\
> }
>
> Is more readable.

I don't care. I'll do it your way (for all 3 patches).

Thanks,

Mathieu

> -- Steve

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