Re: [PATCH tip/core/rcu 1/3] rcu: Provide polling interfaces for Tree RCU grace periods

[Frederic Weisbecker]

On Wed, Mar 03, 2021 at 04:26:30PM -0800, paulmck@xxxxxxxxxx wrote:
> From: "Paul E. McKenney" <paulmck@xxxxxxxxxx>
> 
> There is a need for a non-blocking polling interface for RCU grace
> periods, so this commit supplies start_poll_synchronize_rcu() and
> poll_state_synchronize_rcu() for this purpose.  Note that the existing
> get_state_synchronize_rcu() may be used if future grace periods are
> inevitable (perhaps due to a later call_rcu() invocation).  The new
> start_poll_synchronize_rcu() is to be used if future grace periods
> might not otherwise happen.

By future grace period, you mean if a grace period has been started right
_before_ we start polling, right?


> Finally, poll_state_synchronize_rcu()
> provides a lockless check for a grace period having elapsed since
> the corresponding call to either of the get_state_synchronize_rcu()
> or start_poll_synchronize_rcu().
> 
> As with get_state_synchronize_rcu(), the return value from either
> get_state_synchronize_rcu() or start_poll_synchronize_rcu() is passed in
> to a later call to either poll_state_synchronize_rcu() or the existing
> (might_sleep) cond_synchronize_rcu().
> 
> Signed-off-by: Paul E. McKenney <paulmck@xxxxxxxxxx>
[...]
>  /**
> + * start_poll_state_synchronize_rcu - Snapshot and start RCU grace period
> + *
> + * Returns a cookie that is used by a later call to cond_synchronize_rcu()
> + * or poll_state_synchronize_rcu() to determine whether or not a full
> + * grace period has elapsed in the meantime.  If the needed grace period
> + * is not already slated to start, notifies RCU core of the need for that
> + * grace period.
> + *
> + * Interrupts must be enabled for the case where it is necessary to awaken
> + * the grace-period kthread.
> + */
> +unsigned long start_poll_synchronize_rcu(void)
> +{
> +	unsigned long flags;
> +	unsigned long gp_seq = get_state_synchronize_rcu();
> +	bool needwake;
> +	struct rcu_data *rdp;
> +	struct rcu_node *rnp;
> +
> +	lockdep_assert_irqs_enabled();
> +	local_irq_save(flags);
> +	rdp = this_cpu_ptr(&rcu_data);
> +	rnp = rdp->mynode;
> +	raw_spin_lock_rcu_node(rnp); // irqs already disabled.
> +	needwake = rcu_start_this_gp(rnp, rdp, gp_seq);

I'm a bit surprised we don't start a new grace period instead of snapshotting
the current one.

So if we do this:

   //start grace period gp_num=5

   old = p;
   rcu_assign_pointer(p, new);

   num = start_poll_synchronize_rcu(); // num = 5

   //grace period ends, start new gp_num=6

   poll_state_synchronize_rcu(num); // rcu seq is done

   kfree(old);

Isn't there a risk that other CPUs still see the old pointer?

Of course I know I'm missing something obvious :-)

Thanks.