Re: Q: perf_install_in_context/perf_event_enable are racy?

[Oleg Nesterov <oleg@redhat.com>]

On 01/26, Oleg Nesterov wrote:
>
> > +void task_function_trampoline(void *data)
> > +{
> > +	struct task_function_call *tfc = data;
> > +
> > +	if (this_rq()->curr != tfc->p)
> > +		return;
>
> Yes, I was thinking about checking rq->curr too, but this doesn't
> really help. This closes the race with "prev", but we have the similar
> race with "next".
>
> __perf_install_in_context() should not set ->task_ctx before next
> does perf_event_context_sched_in(). Otherwise it will do nothing,
> it checks cpuctx->task_ctx == ctx.

But of course, we can add rq->in_context_switch or something. This
is more or less equal to Frederic's per-cpu task_events_schedulable
but simpler, because this doesn't depend on perf_task_events.

This is what I had in mind initially but I didn't dare to add the
new member into rq, it is only needed for perf.

Oleg.